Lately I’ve been reading some of George MacDonald Fraser’s Flashman books, which have been quite enjoyable reads in general; I’m reading the books in the order in which the actions in the books supposedly took place, not in the order in which the books were published, and a large number of the words included below are words I encountered in the first three of the books I read (i.e. FlashmanRoyal Flash, and Flashman’s Lady); I decided the post already at that point included a large number of words (the post includes roughly 120 words), so I saw no need to add additional words from the other books in the series in this post as well. I have reviewed a few of the Flashman books I’ve read on goodreads here, here, and here.

Havildar, gimbal, quorum, unmannerly, tribulation, thalassophobia, kiln, sheave, grody, contemn, arcanum, deloping, poulterer, fossorial, catamount, guttersnipe, nabob, frond, matelot, jetty.

Sangar, palliasse, junoesque, cornet, bugle, fettle, toady, thong, trollop, sepoy, wattle, hardtack, snuffle, chunter, ghillie, barker, trousseau, simper, madcap, ramrod.

Welt, landau, declaim, burgomaster, scupper, windlass, maunder, sniffy, sirdar, randy, dowager, toffs, pug, curvet, pish, scriveners, hoyden, manikin, lecher/lechery, busby.

Ruck, leery, ninny, shillyshally, mincing, ringlet, covey, pip, munshi, risaldar, maidan, palankeen/palanquin, forbye, feringhee, cantonment, puggaree, pannikin, dollymop, snook, cordage.

Suet/suety, strumpet, kenspeckle, magsman, scrag, chandler, prigger, chivvy, décolleté, dundrearies, assignation, bruit, purblind, trull, slatterncoffle, doggo, cellarette, cummerbund, agley.

Sampan, wideawake, popsycollation, déshabillé, pinnace, pennant, murk, sprig, linstock, tassel, bangle, trammel, prau, shellback, shako, clobbertaffrail, crinolinetaffeta, commonalty.

April 15, 2017 Posted by | Books, language | Leave a comment


i. “No man is rich enough to buy back his past.” (Oscar Wilde)

ii. “There is a luxury in self-reproach. When we blame ourselves we feel that no one else has a right to blame us.” (-ll-)

iii. “Each new generation asks – What is the meaning of life? A more fertile way of putting the question would be – Why does man need a meaning to life?” (Peter Wessel Zapffe)

iv. “One man’s constant is another man’s variable.” (Alan Perlis)

v. “All scientific work is incomplete – whether it be observational or experimental. All scientific work is liable to be upset or modified by advancing knowledge. That does not confer upon us a freedom to ignore the knowledge we already have, or to postpone the action that it appears to demand at a given time.” (Austin Bradford Hill)

vi. “Most women set out to try to change a man, and when they have changed him they do not like him.” (Marlene Dietrich)

vii. “We speak with our lips to explain, with our throats to convince.” (Malcolm de Chazal)

viii. “Those who do not complain are never pitied.” (Jane Austen)

ix. “It is an aspect of all happiness to suppose that we deserve it.” (Joseph Joubert)

x. “Almost all absurdity of conduct arises from the imitation of those whom we cannot resemble.” (Samuel Johnson)

xi. “Nothing so much prevents our being natural as the desire to seem so.” (Rochefoucauld)

xii. “It is harder to hide feelings we have than to feign those we lack.” (-ll-)

xiii. “Almost all our faults are more pardonable than the methods we resort to to hide them.” (-ll-)

xiv. “Fanaticism consists of redoubling your efforts when you have forgotten your aim.” (George Santayana)

xv. “There is nothing that fear or hope does not make men believe.” (Vauvenargues)

xvi. “There is no rule more invariable than that we are paid for our suspicions by finding what we suspected.” (Henry David Thoreau)

xvii. “It is hard to believe that a man is telling the truth when you know that you would lie if you were in his place.” (H.L. Mencken)

xviii. “Conscience is thoroughly well bred and soon leaves off talking to those who do not wish to hear it.” (Samuel Butler)

xix. “History is not written as it was experienced, nor should it be. The inhabitants of the past know better than we do what it was like to live there, but they were not well placed, most of them, to understand what was happening to them and why.” (Tony Judt)

xx. “Stability is much underappreciated, especially by those who enjoy its benefits.” (Curtis Yarvin)

April 5, 2017 Posted by | Quotes/aphorisms | Leave a comment

A few autism papers

i. The anterior insula in autism: Under-connected and under-examined.

“While the past decade has witnessed a proliferation of neuroimaging studies of autism, theoretical approaches for understanding systems-level brain abnormalities remain poorly developed. We propose a novel anterior insula-based systems-level model for investigating the neural basis of autism, synthesizing recent advances in brain network functional connectivity with converging evidence from neuroimaging studies in autism. The anterior insula is involved in interoceptive, affective and empathic processes, and emerging evidence suggests it is part of a “salience network” integrating external sensory stimuli with internal states. Network analysis indicates that the anterior insula is uniquely positioned as a hub mediating interactions between large-scale networks involved in externally- and internally-oriented cognitive processing. A recent meta-analysis identifies the anterior insula as a consistent locus of hypoactivity in autism. We suggest that dysfunctional anterior insula connectivity plays an important role in autism. […]

Increasing evidence for abnormal brain connectivity in autism comes from studies using functional connectivity measures […] These findings support the hypothesis that under-connectivity between specific brain regions is a characteristic feature of ASD. To date, however, few studies have examined functional connectivity within and between key large-scale canonical brain networks in autism […] The majority of published studies to date have examined connectivity of specific individual brain regions, without a broader theoretically driven systems-level approach.

We propose that a systems-level approach is critical for understanding the neurobiology of autism, and that the anterior insula is a key node in coordinating brain network interactions, due to its unique anatomy, location, function, and connectivity.”

ii. Romantic Relationships and Relationship Satisfaction Among Adults With Asperger Syndrome and High‐Functioning Autism.

“Participants, 31 recruited via an outpatient clinic and 198 via an online survey, were asked to answer a number of self-report questionnaires. The total sample comprised 229 high-functioning adults with ASD (40% males, average age: 35 years). […] Of the total sample, 73% indicated romantic relationship experience and only 7% had no desire to be in a romantic relationship. ASD individuals whose partner was also on the autism spectrum were significantly more satisfied with their relationship than those with neurotypical partners. Severity of autism, schizoid symptoms, empathy skills, and need for social support were not correlated with relationship status. […] Our findings indicate that the vast majority of high-functioning adults with ASD are interested in romantic relationships.”

Those results are very different from other results in the field – for example: “[a] meta-analysis of follow-up studies examining outcomes of ASD individuals revealed that, [o]n average only 14% of the individuals included in the reviewed studies were married or ha[d] a long-term, intimate relationship (Howlin, 2012)” – and one major reason is that they only include high-functioning autistics. I feel sort of iffy about the validity of the selection method used for procuring the online sample, this may also be a major factor (almost one third of them had a university degree so this is definitely not a random sample of high-functioning autistics; ‘high-functioning’ autistics are not that high-functioning in the general setting. Also, the sex ratio is very skewed as 60% of the participants in the study were female. A sex ratio like that may not sound like a big problem, but it is a major problem because a substantial majority of individuals with mild autism are males. Whereas the sex ratio is almost equal in the context of syndromic ASD, non-syndromic ASD is much more prevalent in males, with sex ratios approaching 1:7 in milder cases (link). These people are definitely looking at the milder cases, which means that a sample which skews female will not be remotely similar to most random samples of such individuals taken in the community setting. And this matters because females do better than males. A discussion can be had about to which extent women are under-diagnosed, but I have not seen data convincing me this is a major problem. It’s important to keep in mind in that context that the autism diagnosis is not based on phenotype alone, but on a phenotype-environment interaction; if you have what might be termed ‘an autistic phenotype’ but you are not suffering any significant ill effects as a result of this because you’re able to compensate relatively well (i.e. you are able to handle ‘the environment’ reasonably well despite the neurological makeup you’ve ended up with), you should not get an autism diagnosis – a diagnostic requirement is ‘clinically significant impairment in functioning’.

Anyway some more related data from the publication:

“Studies that analyze outcomes exclusively for ASD adults without intellectual impairment are rare. […] Engström, Ekström, and Emilsson (2003) recruited previous patients with an ASD diagnosis from four psychiatric clinics in Sweden. They reported that 5 (31%) of 16 adults with ASD had ”some form of relation with a partner.” Hofvander et al. (2009) analyzed data from 122 participants who had been referred to outpatient clinics for autism diagnosis. They found that 19 (16%) of all participants had lived in a long-term relationship.
Renty and Roeyers (2006) […] reported that at the time of the[ir] study 19% of 58 ASD adults had a romantic relationship and 8.6% were married or living with a partner. Cederlund, Hagberg, Billstedt, Gillberg, and Gillberg (2008) conducted a follow-up study of male individuals (aged 16–36 years) who had been diagnosed with Asperger syndrome at least 5 years before. […] at the time of the study, three (4%) [out of 76 male ASD individuals] of them were living in a long-term romantic relationship and 10 (13%) had had romantic relationships in the past.”

A few more data and observations from the study:

“A total of 166 (73%) of the 229 participants endorsed currently being in a romantic relationship or having a history of being in a relationship; 100 (44%) reported current involvement in a romantic relationship; 66 (29%) endorsed that they were currently single but have a history of involvement in a romantic relationship; and 63 (27%) participants did not have any experience with romantic relationships. […] Participants without any romantic relationship experience were significantly more likely to be male […] According to participants’ self-report, one fifth (20%) of the 100 participants who were currently involved in a romantic relationship were with an ASD partner. […] Of the participants who were currently single, 65% said that contact with another person was too exhausting for them, 61% were afraid that they would not be able to fulfil the expectations of a romantic partner, and 57% said that they did not know how they could find and get involved with a partner; and 50% stated that they did not know how a romantic relationship works or how they would be expected to behave in a romantic relationship”

“[P]revious studies that exclusively examined adults with ASD without intellectual impairment reported lower levels of romantic relationship experience than the current study, with numbers varying between 16% and 31% […] The results of our study can be best compared with the results of Hofvander et al. (2009) and Renty and Roeyers (2006): They selected their samples […] using methods that are comparable to ours. Hofvander et al. (2009) found that 16% of their participants have had romantic relationship experience in the past, compared to 29% in our sample; and Renty and Roeyers (2006) report that 28% of their participants were either married or engaged in a romantic relationship at the time of their study, compared to 44% in our study. […] Compared to typically developed individuals the percentage of ASD individuals with a romantic relationship partner is relatively low (Weimann, 2010). In the group aged 27–59 years, 68% of German males live together with a partner, 27% are single, and 5% still live with their parents. In the same age group, 73% of all females live with a partner, 26% live on their own, and 2% still live with their parents.”

“As our results show, it is not the case that male ASD individuals do not feel a need for romantic relationships. In fact, the contrary is true. Single males had a greater desire to be in a romantic relationship than single females, and males were more distressed than females about not being in a romantic relationship.” (…maybe in part because the females who were single were more likely than the males who were single to be single by choice?)

“Our findings showed that being with a partner who also has an ASD diagnosis makes a romantic relationship more satisfying for ASD individuals. None of the participants, who had been with a partner in the past but then separated, had been together with an ASD partner. This might indicate that once a person with ASD has found a partner who is also on the spectrum, a relationship might be very stable and long lasting.”

Reward Processing in Autism.

“The social motivation hypothesis of autism posits that infants with autism do not experience social stimuli as rewarding, thereby leading to a cascade of potentially negative consequences for later development. […] Here we use functional magnetic resonance imaging to examine social and monetary rewarded implicit learning in children with and without autism spectrum disorders (ASD). Sixteen males with ASD and sixteen age- and IQ-matched typically developing (TD) males were scanned while performing two versions of a rewarded implicit learning task. In addition to examining responses to reward, we investigated the neural circuitry supporting rewarded learning and the relationship between these factors and social development. We found diminished neural responses to both social and monetary rewards in ASD, with a pronounced reduction in response to social rewards (SR). […] Moreover, we show a relationship between ventral striatum activity and social reciprocity in TD children. Together, these data support the hypothesis that children with ASD have diminished neural responses to SR, and that this deficit relates to social learning impairments. […] When we examined the general neural response to monetary and social reward events, we discovered that only TD children showed VS [ventral striatum] activity for both reward types, whereas ASD children did not demonstrate a significant response to either monetary or SR. However, significant between-group differences were shown only for SR, suggesting that children with ASD may be specifically impaired on processing SR.”

I’m not quite sure I buy that the methodology captures what it is supposed to capture (“The SR feedback consisted of a picture of a smiling woman with the words “That’s Right!” in green text for correct trials and a picture of the same woman with a sad face along with the words “That’s Wrong” in red text for incorrect trials”) (this is supposed to be the ‘social reward feedback’), but on the other hand: “The chosen reward stimuli, faces and coins, are consistent with those used in previous studies of reward processing” (so either multiple studies are of dubious quality, or this kind of method actually ‘works’ – but I don’t know enough about the field to tell which of the two conclusions apply).

iv. The Social Motivation Theory of Autism.

“The idea that social motivation deficits play a central role in Autism Spectrum Disorders (ASD) has recently gained increased interest. This constitutes a shift in autism research, which has traditionally focused more intensely on cognitive impairments, such as Theory of Mind deficits or executive dysfunction, while granting comparatively less attention to motivational factors. This review delineates the concept of social motivation and capitalizes on recent findings in several research areas to provide an integrated picture of social motivation at the behavioral, biological and evolutionary levels. We conclude that ASD can be construed as an extreme case of diminished social motivation and, as such, provides a powerful model to understand humans’ intrinsic drive to seek acceptance and avoid rejection.”

v. Stalking, and Social and Romantic Functioning Among Adolescents and Adults with Autism Spectrum Disorder.

“We examine the nature and predictors of social and romantic functioning in adolescents and adults with ASD. Parental reports were obtained for 25 ASD adolescents and adults (13-36 years), and 38 typical adolescents and adults (13-30 years). The ASD group relied less upon peers and friends for social (OR = 52.16, p < .01) and romantic learning (OR = 38.25, p < .01). Individuals with ASD were more likely to engage in inappropriate courting behaviours (χ2 df = 19 = 3168.74, p < .001) and were more likely to focus their attention upon celebrities, strangers, colleagues, and ex-partners (χ2 df = 5 =2335.40, p < .001), and to pursue their target longer than controls (t = -2.23, df = 18.79, p < .05).”

“Examination of relationships the individuals were reported to have had with the target of their social or romantic interest, indicated that ASD adolescents and adults sought to initiate fewer social and romantic relationships but across a wider variety of people, such as strangers, colleagues, acquaintances, friends, ex-partners, and celebrities. […] typically developing peers […] were more likely to target colleagues, acquaintances, friends, and ex-partners in their relationship attempts, whilst the ASD group targeted these less frequently than expected, and attempted to initiate relationships significantly more frequently than is typical, with strangers and celebrities. […] In attempting to pursue and initiate social and romantic relationships, the ASD group were reported to display a much wider variety of courtship behaviours than the typical group. […] ASD adolescents and adults were more likely to touch the person of interest inappropriately, believe that the target must reciprocate their feelings, show obsessional interest, make inappropriate comments, monitor the person’s activities, follow them, pursue them in a threatening manner, make threats against the person, and threaten self-harm. ASD individuals displayed the majority of the behaviours indiscriminately across all types of targets. […] ASD adolescents and adults were also found […] to persist in their relationship pursuits for significantly longer periods of time than typical adolescents and adults when they received a negative or no response from the person or their family.”

April 4, 2017 Posted by | autism, Neurology, Papers, Psychology | Leave a comment

Health econ stuff

In a post I published a few weeks ago I mentioned that I had decided against including some comments and observations I had written about health economics in that post because the post was growing unwieldy, but that I might post that stuff later on in a separate post. This post will include those observations, as well as some additional details I added to the post later. This sort of post is the sort of post that usually does not get past the ‘draft’ stage (in wordpress you can save posts you intend to publish later on as drafts), and as is usually the case for posts like these I already regret having written it, for multiple reasons. I should warn you from the start that this post is very long and will probably take you some time to read.

Anyway, the starting point for this post was some comments related to health insurance and health economics which I left on SSC in the past. A lot more people read those comments on SSC than will read this post so the motivation for posting it here was not to ‘increase awareness’ of the ideas and observations included in some kind of general sense; my primary motivation for adding this stuff here is rather that it’s a lot easier for me personally to find stuff I’ve written when it’s located here on this blog rather than elsewhere on the internet, and I figure that some of the things I wrote back then are topics which might well come up again later, and it would be convenient for me in that case to have a link at hand. Relatedly I have added many additional comments and observations in this post not included in the primary exchange, which it is no longer possible for me to do on SSC as my comments are no longer editable on that site.

Although the starting point for the post was as mentioned a comment exchange, I decided early on against just ‘quoting myself’ in this post, and I have thus made some changes in wording and structure in order to increase the precision of the statements included and in order to add a bit of context making the observations below easier to read and understand (and harder to misread). Major topics to which the observations included in this post relate are preventable diseases, the level of complexity that is present in the health care sector, and various topics which relate to health care cost growth. Included in the post are some perhaps not sufficiently well known complications which may arise in the context of the discussion of how different financing schemes may relate to various outcomes, and to cost growth. Much of the stuff included will probably be review to people who’ve read my previous posts on health economics, but that’s to be expected considering the nature of this post.

Although ‘normative stuff’ is not what interests me most – I generally tend to prefer discussions where the aim is to identify what happens if you do X, and I’ll often be happy to leave the discussion of whether outcome X or Y is ‘best’ to others – I do want to start out with stating a policy preference, as this preference was the starting point for the aforementioned debate that lead to the origination of this post. At the outset I should thus make clear that I would in general be in favour of changes to the financial structure of health care systems where people who take avoidable risks which systematically and demonstrably increase their expected health care expenditures at the population level pay a larger proportion of the cost than do people who did not take such avoidable risks.

Most developed societies have health care systems which are designed in a way that implicitly to some extent subsidize unhealthy behaviours. An important note in this context is incidentally that one way of looking at these things is that if you are not explicitly demanding people who behave in risky ways which tend to increase their expected costs to pay more for their health care (/insurance), then you are in fact by virtue of not doing this implicitly subsidizing those unhealthy individuals/behaviours. I mention this because some people might not like the idea of ‘subsidizing healthy behaviours’ (‘health fascism’) – which from a certain point of view is what you do if you charge people who behave in unhealthy ways more. Maybe some people would take issue with words like ‘subsidy’ and ‘implicit’, but regardless of what you call these things the major point that is important to have in mind here is that if one group of people (e.g. ‘unhealthy people’) cost more to treat (/are ill more often, get illnesses related to their behaviours, etc., etc.) than another group of people (‘healthy people’), then if you need to finance this shortfall – which you do, as you face a budget constraint – there are only two basic ways to do this; you can either charge the high-cost group (‘unhealthy people’) more, or you can require the other group (‘healthy people’) to make up the difference. Any scheme which deals with such a case of unequal net contribution rates are equivalent either to one of those schemes or a mix of the two, regardless of what you call things and how it’s done, and regardless of which groups we are talking about (old people also have higher health care expenditures than do young people, and most health care systems implicitly redistribute income from the young to the old). If you’re worried about ‘health fascism’ and the implications of subsidizing healthy behaviours (/’punishing’ unhealthy behaviours) you should at least keep in mind that if the health care costs of people who live healthy lives and people who do not are dissimilar then any system that deals with this issue – which all systems must – can either choose to ‘subsidize’ healthy behaviours or unhealthy behaviours; there’s no feasible way to design a ‘neutral system’ if the costs of the groups are dissimilar.

Having said all this, the very important next point is then that it is much more difficult to make simple schemes that would accomplish an outcome in which people who engage in unhealthy behaviours are required to pay more without at the same time introducing a significant number of new problems than people who are not familiar with this field would probably think it is. And it’s almost certainly much harder to evaluate if the proposed change actually accomplished what you wanted to accomplish than you think it is. Even if we are clear about what we want to accomplish and can all agree that that is what we are aiming for – i.e. we are disregarding the political preferences of large groups of voters and whether the setup in question is at all feasible to accomplish – this stuff is really much harder than it looks, for many reasons.

Let’s start out by assuming that smoking increases the risk of disease X by 50%. Say you can’t say which of the cases of X are caused by smoking, all you know is that smoking increases the risk at the population level. Say you don’t cover disease X at all if someone smokes, that is, smokers are required to pay the full treatment cost out of pocket if they contract disease X. It’s probably not too controversial to state that this approach might by some people be perceived of as not completely ‘fair’ to the many smokers who would have got disease X even if they had not smoked (a majority in this particular case, though of course the proportion will vary with the conditions and the risk factors in question). Now, a lot of the excess health care costs related to smoking are of this kind, and it is actually a pretty standard pattern in general with risk factors – smoking, alcohol, physical inactivity, etc. You know that these behaviours increase risk, but you usually can’t say for certain which of the specific cases you observe in clinical practice are actually (‘perfectly’/’completely’/’partially’?) attributable to the behaviour. And quite often the risk increase associated with a specific behaviour is actually really somewhat modest, compared to the relevant base rates, meaning that many of the people who engage in behaviours which increase risk and who get sick might well have got sick even if they hadn’t engaged in those risky behaviours.

On top of this problem usually it’s also the case that risk factors interact with each other. Smoking increases the risk of cancer of the esophagus, but so does alcohol and obesity, and if a person both smokes and drinks the potential interaction effect may not be linear – so you most likely often can’t just identify individual risk factors in specific studies and then pool them later and add them all together to get a proper risk assessment. A further complication is that behaviours may both increase as well as decrease risk – to stick with the example, diets high in fruits and vegetables both lower the risk of cancer of the esophagus. Exercise probably does as well – we know that exercise has important and highly complex effects on immune system function (see e.g. this post). Usually a large number of potential risk factors is at play at the same time, there may be multiple variables which lower risk and are also important to include if you want a proper risk assessment, and even if you knew in theory which interaction terms were likely to be relevant, you might even so find yourself in a situation unable to estimate the interaction terms of interest – this might take high-powered studies with large numbers of patients, which may not be available or the results of such high-powered studies may not apply to your specific subgroup of patients. Cost-effectiveness is also an issue – it’s expensive to assess risk properly. One take-away is that you’ll still have a lot of unfairness in a modified contribution rate model, and even evaluating fairness aspects of the change may be difficult to impossible because to some extent this question is unknowable. You might find yourself in a situation where you charge the obese guy more because obesity means he’s high risk, but in reality he is actually lower risk than is the non-fat guy who is charged a lower rate, because he also exercises and eats a lot of fruits and vegetables, which the other guy doesn’t.

Of course the above paragraph took it for granted that it was even possible to quantify the excess costs attributable to a specific condition. That may not be easy at all to do, and there may be large uncertainties involved. The estimated excess cost will depend upon a variety of factors which may or may not be of interest to the party performing the analysis, for example it may be very important which time frame you’re looking at and which discounting methodology is applied (see e.g. the last paragraph in this post). The usual average vs marginal cost problem (see the third-last paragraph in the post to which I link in the previous sentence – this post also has more on this topic) also applies here and is related to ‘the fat guy who exercises and is low-risk’-problem; ideally you’d want to charge people with higher health care utilization levels more (again, in a setting where we assume the excess cost is associated with life-style variables which are modifiable – this was our starting point), but if there’s a large amount of variation in costs across individuals in the specific subgroups of interest and you only have access to average costs rather than individual-level costs, then a scheme only taking into account the differences in the averages may be very sub-optimal when you look at it from the viewpoint of the individual. Care needs to be taken to avoid problems like e.g. Simpson’s paradox.

Risk factors are not the only things that cluster; so do diseases. An example:

“An analysis of the Robert Koch-Institute (RKI) from 2012 shows that more than 50 % of German people over 65 years suffer from at least one chronic disease, approximately 50 % suffer from two to four chronic diseases, and over a quarter suffer from five or more diseases [3].” (link)

78.3 % of the type 2 diabetics also suffered from hypertension in that study. Does this fact make it easier or harder to figure out what is ‘the true cost contribution’ of ‘type 2 diabetes’ and ‘hypertension’ (and, what we’re ultimately interested in in this setting – the ‘true cost contribution’ of the unhealthy behaviours which lead some individuals to develop type 2 diabetics and hypertension who would not otherwise have developed diabetes and/or hypertension (…/as early as they did)? It should be noted that diabetes was estimated to account for 11 % of total global healthcare expenditure on adults in 2013 (link). That already large proportion is expected to rise substantially in the decades to come – if you’re interested in cost growth trajectories, this is a major variable to account for. Attributability is really tricky here, and perhaps even more tricky in the case of hypertension – but for what it’s worth, according to a CDC estimate hypertension cost the US $46 billion per year, or ~$150/per person per year.

Anyway, you look at the data and you make guesses, but the point is that doctor Smith won’t know for certain if Mr. Hanson would have had a stroke even if he hadn’t smoked or not. A proposal of not providing payment for a health care service or medical product in the case of an ‘obviously risky-behaviour-related-health-condition’ may sometimes appear to be an appealing proposition and you sometimes see people make this sort of proposal in discussions of this nature, but it tends to be very difficult when you look at the details to figure out just what those ‘obviously risky-behaviour-related-health-conditions’ are, and even harder to make even remotely actuarially fair adjustments to the premiums and coverage patterns to reflect the risk. Smoking and lung cancer is a common example of a relatively ‘clean’ case, but most cases are ‘less clean’ and even here there are complications; a substantial proportion of lung cancer cases are not caused by tobacco – occupational exposures also cause a substantial proportion of cases, and: “If considered in its own disease category […] lung cancer in never smokers would represent the seventh leading cause of cancer mortality globally, surpassing cancers of the cervix, pancreas, and prostate,5 and among the top 10 causes of death in the United States.” (link) Occupational exposures (e.g. asbestos) are not likely to fully account for all cases, and for example it has also been found that other variables, including previous pneumonia infections and tuberculosis, affect risk (here are a couple of relevant links to some previous coverage I wrote on these topics).

I think many people who have preferences of this nature (‘if it’s their own fault they’re sick, they should pay for it themselves’) underestimate how difficult it may be to make changes which could be known with a reasonable level of certainty to actually have the intended consequences, even assuming everybody agreed on the goal to be achieved. This is in part because there are many other aspects and complications which need to be addressed as well. Withholding payment in the case of costly preventative illness may for example in some contexts increase cost, rather than decrease them. The risk of complications of some diseases – an important cost driver in the context of diabetes – tends to be dependent on post-diagnosis behavioural patterns. The risk of developing diabetes complications will depend upon the level of glycemic control. If you say you won’t cover complications at all in the case of ‘self-inflicted disease X’, then you also to some extent tend to remove the option of designing insurance schemes which might lower cost and complication rates post-diagnosis by rewarding ‘good’ (risk-minimizing) behaviours post-diagnosis and punishing ‘bad’ (risk-increasing) behaviours. This is not desirable in the context of diseases where post-diagnosis behaviour is an important component of the cost function, as it certainly is in the diabetes context. There are multiple potential mechanisms here, some of which are disease specific (e.g. suboptimal diet in a diagnosed type 2 diabetic) and some of which may not be (a more general mechanism could e.g. be lowered compliance/adherence to treatment in the uncovered populations because they can’t afford the drugs which are required to treat their illness; though the cost-compliance link is admittedly not completely clear in the general case, there are certainly multiple diseases where lowered compliance to treatment would be expected to increase cost long-term).

And again, also in the context of complications fairness issues are not as simple to evaluate as people might like them to be; some people may have a much harder time controlling their disease than others, or they may be more susceptible to complications given the same behaviour. Some may already have developed complications by the time of diagnosis. Such issues make it difficult to design simple rules which would achieve what you want them to achieve without having unfortunate side-effects; for example a rule that a microvascular diabetes-related complication is automatically ‘your own fault’ (so we won’t pay for it), which might be motivated by the substantial amount of research linking glycemic control with complication risk, would punish some diabetics who have had the disease for a longer amount of time (many complications are not only strongly linked to Hba1c but also display a substantial degree of duration-dependence; for example in type 1 diabetics one study found that diabetic retinopathy was present in 13% of patients with a duration of disease less than 5 years, whereas the corresponding figure was 90% for individuals with a disease duration of 10–15 years (Sperling et al., p. 393). I also recall reading a study finding that Hba1c itself is increasing with diabetes duration, which may be partly accounted for by the higher risk of hypoglycemia related to hypoglycemia-unawareness-syndromes in individuals with long-standing disease), individuals with diseases which are relatively hard to control (perhaps due to genetics, or maybe again due to the fact that they have had the disease for a longer amount of time; the presence of hypoglycemia unawareness is as alluded to above to a substantial degree duration-dependent, and this problem increases the risk of hospitalizations, which are expensive), diabetics who developed complications before they knew they were sick (a substantial proportion of type 2 diabetics develop some degree of microvascular damage pre-diagnosis), and diabetics with genetic variants which confer an elevated risk of complications (“observations suggest that involvement of genetic factors is increasing the risk of complications” (Sperling et al., p. 226), and for example in the DCCT trial familial clustering of both neuropathy and retinopathy was found; clustering which persisted after controlling for Hba1c – for more on these topics, see e.g. Sperling et al.’s chapter 11).

Other decision rules would similarly lead to potentially problematic incentives and fairness issues; for example requiring individuals to meet a specific Hba1c goal might be more desirable than to just not cover complications, but that one also leads to potential problems; ideally such an Hba1c goal should be individualized, because of the above-mentioned complexities and others I have not mentioned here; to require a newly-diagnosed individual to meet the same goals as someone who has had diabetes for decades does not make sense, and neither does it make sense to require these two groups to meet exactly the same Hba1c goal as the middle-aged female diabetic who desires to become pregnant (diabetes greatly increases the risk of pregnancy complications, and strict glycemic control is extremely important in this patient group). It’s important to note that these issues don’t just relate to whether or not the setup is perceived of as fair, but it also relates to whether or not you would expect the intended goals to actually be met or not when you implement the rule. If you were to require that a long-standing diabetic with severe hypoglycemia unawareness had to meet the same Hba1c goal as the newly diagnosed individual, this might well lead to higher overall cost, because said individual might suffer a large number of hypoglycemia-related hospitalizations which would have been avoidable if a more lax requirement was imposed; when you decrease Hba1c you decrease the risk of long-term complications, but you increase the risk of hypoglycemia. A few numbers might make it easier to make sense of how expensive hospitalizations really are, and why I emphasize them here. In this diabetes-care publication they assign a cost for an inpatient day for a diabetes-related hospitalization at $2,359 and an emergency visit at ~$800. The same publication estimates the total average annual excess expenditures of diabetics below the age of 45 at $4,394. Going to the hospital is really expensive (43% of the total medical costs of diabetes are accounted for by hospital inpatient care in that publication).

A topic which was brought up in the SSC discussion was the question of the extent to which private providers have a greater incentive to ‘get things right’ in terms of assessing risk. I don’t take issue with this notion in general, but there are a lot of complicating factors in the health care context. One factor of interest is that it is costly to get things right. If you’re looking at this from an insurance perspective, larger insurance providers may be better at getting things right because they can afford to hire specialists who provide good cost estimates – getting good cost estimates is really hard, as I’ve noted above. Larger providers translate into fewer firms, which increases firm concentration and may thus increase collusion risk, which may again increase the prices of health care services. Interestingly if your aim is to minimize health care cost growth increased market power of private firms may actually be a desirable state of affairs/goal, because cost growth is a function of both unit prices and utilization levels, and higher premiums are likely to translate into lower utilization rates, which may lower overall costs and -cost growth. I decided to include this observation here also in order to illustrate that what is an optimal outcome depends on what your goal is, and in the setting of the health care sector you sometimes need to be very careful about thinking about what your actual goal is, and which other goals might be relevant.

When private insurance providers become active in a market that also includes a government entity providing a level of guaranteed coverage, total medical outlays may easily increase rather than decrease. The firms may meed an unmet need, but some of that unmet need may be induced demand (here’s a related link). Additionally, the bargaining power of various groups of medical personnel may change in such a setting, leading to changes in compensation schedules which may not be considered desirable/fair. An increase in total outlays may or may not be considered a desirable outcome, but this does illustrate once again the point that you need to be careful about what you are trying to achieve.

There’s a significant literature on how the level of health care integration, both at the vertical and horizontal level, both in terms of financial structure and e.g. in terms of service provision structure, may impact health care costs, and this is an active area of research where we in some contexts do not yet know the answers.

Even when cost minimization mechanisms are employed in the context of private firms and the firm in question is efficient, the firm may not internalize all relevant costs. This may paradoxically lead to higher overall cost, due to coverage decisions taken ‘upstream’ influencing costs ‘downstream’ in an adverse manner; I have talked about this topic on this blog before. A diabetic might be denied coverage of glucose testing materials by his private insurer, and that might mean that the diabetic instead gets hospitalized for a foreseeable and avoidable complication (hypoglycemic coma due to misdosing), but because it might not be the same people paying for the testing material and the subsequent hospitalization it might not matter to the people denying coverage of the testing materials, and/so they won’t take it into account when they’re making their coverage decisions. That sort of thing is quite common in the health care sector – different entities pay for and receive payments for different things, and this is once again a problem to keep in mind if you’re interested in health care evaluation; interventions which seem to lower cost may not do so in reality, because the intervention lead to higher health care utilization elsewhere in the system. If incentives are not well-aligned things may go badly wrong, and they are often not well-aligned in the health care sector. When both the private and public sectors are involved in either the financial arrangements and/or actual health service provision – which is the default health care system setup for developed societies – this usually leads to highly complex systems, where the scope for such problems to appear seems magnified, rather than the opposite. I would assume that in many cases it matters a lot more that incentives are well-aligned than which specific entity is providing insurance or health care in the specific context, in part a conclusion drawn from the coverage included in Simmons, Wenzel & Zgibor‘s book.

In terms of the incentive structures of the people involved in the health care sector, this stuff is of course also adding another layer of complexity. In all sectors of the economy you have people with different interests who interact with each other, and when incentives change outcomes change. Outcomes may be car batteries, or baseball bats, or lectures. Evaluating outcomes is easier in some settings than in others, and I have already touched upon some of the problems that might be present when you’re trying to evaluate outcomes in the health care context. How easy it is to evaluate outcomes will naturally vary across sub-sectors of the health care sector but a general problem which tends to surface here is the existence of various forms of asymmetrical information. There are multiple layers, but a few examples are worth mentioning. To put it bluntly, the patient tends to know his symptoms and behavioural patterns – which may be disease-relevant, and this aspect is certainly important to include when discussing preventative illnesses caused at least in part by behaviours which increase the risk of said illnesses – better than his doctor, and the doctor will in general tend to know much more about the health condition and potential treatment options than will the patient. The patient wants to get better, but he also wants to look good in the eyes of the doctor, which means he might not be completely truthful when interacting with the doctor; he might downplay how much alcohol he drinks, misrepresent how often he exercises, or he may lie about smoking habits or about how much he weighs. These things make risk-assessments more difficult than they otherwise might have been. As for the GPs, usually we here have some level of regulation which restricts their behaviour to some extent, and part of the motivation for such regulation is to reduce the level of induced demand which might otherwise be the result of information asymmetry in the context of stuff like relevant treatment effects. If a patient is not sufficiently competent to evaluate the treatments he receives (‘did the drug the doctor ordered really work, or would I have gotten better without it?’), there’s a risk he might be talked into undergoing needless procedures or take medications for which he has no need, especially if the doctor who advises him has a financial interest in the treatment modality on offer.

General physicians have different incentives from nurses and specialists working in hospitals, and all of these groups may experience conflicts of interests when they’re dealing with insurance providers and with each other. Patients as mentioned have their own set of incentives, which may not align perfectly with those of the health care providers. Different approaches to how to deal with such problems lead to different organizational setups, all of which influence both the quantity and quality of care, subject to various constraints. It’s an active area of research whether decreasing competition between stakeholders/service providers may decrease costs; one thing that is relatively clear from diabetes research with which I have familiarized myself is that when different types of care providers coordinate activities, this tends to lead to better outcomes (and sometimes, but not always, lower costs), because some of the externalized costs become internalized by virtue of the coordination. It seems very likely to me that conclusions to such questions will be different for different subsectors of the health care sector. A general point might be that more complex diseases should be expected to be more likely to generate cost savings from increased coordination than should relatively simple diseases (if you’re fuzzy about what the concept of disease complexity refers to, this post includes some relevant observations). This may be important, because complex diseases also should probably tend to be more expensive to treat in general, because the level of need in patients is higher.

It’s perhaps hardly surprising, considering the problems I’ve already discussed related to how difficult it may be to properly assess costs, that there’s a big discussion to be had about how to even estimate costs (and benefits) in specific contexts, and that people write books about these kinds of things. A lot of things have already been said on this topic and a lot more could be said, but one general point perhaps worth repeating is that it may in the health care sector be very difficult to figure out what things (‘truly’) cost (/’is worth’). If you only have a public sector entity dealing with a specific health problem and patients are not charged for receiving treatment, it may be very difficult to figure out what things ‘should’ cost because relevant prices are simply missing from the picture. You know what the government entity paid the doctors in wages and what it paid for the drugs, but the link between payment and value is sometimes a bit iffy here. There are ways to at least try to address some of these issues, but as already noted people write books about these kinds of things so I’m not going to provide all the highlights here – I refer to the previous posts I’ve written on these topics instead.

Another important related point is that medical expenditures and medical costs are not synonyms. There are many costs associated with illness which are not directly related to e.g. a payment to a doctor. People who are ill may be less productive while they are at work, they may have more sick-days, they may retire earlier, their spouse may cut down on work hours to take care of them instead of going to work, a family caretaker may become ill as a result of the demands imposed by the caretaker role (for example Alzheimer’s disease significantly increases the risk of depression in the spouse). Those costs are relevant, there are literatures on these things, and in some contexts such ‘indirect costs’ (e.g. lower productivity at work and early retirement) may make up a very substantial proportion of the total costs of a health condition. I have seen diabetes cost estimates which indicated that the indirect costs may account for as much as 50 % of the total costs.

If there’s a significant disconnect between total costs and medical expenditures then minimizing expenditures may not be desirable from an economic viewpoint. A reasonable assessment model will/should in the context of models of outlays include both a monetary cost parameter and a quality/quantity (ideally both) parameter; if you neglect to take account of the latter, in some sense you’re only dealing with what you pay out, not what you get for that payment (which is relevant). If you don’t take into account indirect costs you implicitly allow cost switching practices to potentially muddle the picture and make assessments more difficult; for example if you provide fewer long-term care facilities then the number of people involved in ‘informal care’ (e.g. family members having to take care of granny) will go up, and that is going to have secondary effects downstream which should also be assessed (you improve the budget in the context of the long-term care facilities, but you may at the same time increase demands on e.g. psychiatric institutions and marginally lower especially the female labour market participation rate. The net effect may still be positive, but the point is that an evaluation will/should include costs like these in the analysis, at least if you want anything remotely close to the full picture).

Let’s return to those smokers we talked about earlier. A general point not mentioned yet is that if you don’t cover smokers in the public sector because of cost considerations, many of them may also not be covered by private insurance either. This is because a group of individuals that is high risk and expensive to treat will be demanded high premiums (or the insurance providers would go out of business), and for the sake of this discussion we’re now assuming smokers are expensive. If that is so, many of them probably would not be able to afford the premiums demanded. Now, one of the health problems which are very common in smokers is chronic obstructive pulmonary disease (COPD). Admission rates for COPD patients differ as much as 10-fold between European countries, and one of the most important parameters regarding pharmacoeconomics is the hospitalization rate (both observations are from this text). What does this mean? It means that we know that admission rate from COPD is highly responsive to the treatment regime; populations well-treated have much fewer hospitalizations. 4% of all Polish hospitalizations are due to COPD. If you remove the public sector subsidies, the most likely scenario you get seems to me to be a poor-outcomes scenario with lots of hospitalizations. Paying for those is likely to be a lot more expensive than it is to treat the COPD pharmacologically in the community. And if smokers aren’t going to be paying for it, someone else will have to do that. If you both deny them health insurance and refuse them treatment if they cannot pay for it they may just die of course, but in most cost-assessment models that’s a high-cost outcome, not a low-cost outcome (e.g. due to lost work-life productivity etc. Half of people with COPD are of working age, see the text referred to above.). This is one example where the ‘more fair’ option might lead to higher costs, rather than lower costs. Some people might still consider such an outcome desirable, it depends on the maximand of interest, but such outcomes are worth considering when assessing the desirability of different systems.

A broadly similar dynamic, in the context of post-diagnosis behaviour and links to complications and costs, may be present in the context of type 2 diabetes. I know much more about diabetes than I do about respirology, but certainly in the case of diabetes this is a potentially really big problem. Diabetics who are poorly regulated tend to die a lot sooner than other people, they develop horrible complications, they stop being able to work, etc. etc. Some of those costs you can ignore if you’re willing to ‘let them die in the streets’ (as the expression goes), but a lot of those costs are indirect costs due to lower productivity, and those costs aren’t going anywhere, regardless of who may or may not be paying the medical bills of these people. Even if they have become sick due to a high-risk behaviour of their own choosing, their health care costs post-diagnosis will still be highly dependent upon their future medical care and future health insurance coverage. Denying them coverage for all diabetes-related costs post-diagnosis may, paradoxical though it may seem to some, not be the cost-minimizing option.

I already talked about information asymmetries. Another problematic aspect linked to information management also presents itself here in a model of this nature (‘deny all diabetes-related coverage to known diabetics’); people who suspect they might be having type 2 diabetes may choose not to disclose this fact to a health care provider because of the insurance aspect (denial of coverage problems). Insurance providers can of course (and will try to) counter this by things like mandatory screening protocols, but this is expensive, and even assuming they are successful you again not only potentially neglect to try to minimize the costs of the high-cost individuals in the population (the known diabetics, who might be cheaper long-term if they had some coverage), you also price a lot of non-diabetics out of the market (because premiums went up to pay for the screening). And some of those non-diabetics are diabetics to-be, who may get a delayed diagnosis as a result, with an associated higher risk of (expensive) complications. Again, as in the smoking context if the private insurer does not cover the high-cost outcomes someone else will have to do that, and the blind diabetic in a wheel-chair is not likely to be able to pay for his dialysis himself.

More information may in some situations lead to a breakdown in insurance markets. This is particularly relevant in the context of genetics and genetic tests. If you have full information, or close to it, the problem you have to some extent stops being an insurance problem and instead becomes a problem of whether or not to, and to which extent you want to-, explicitly compensate people for having been dealt a bad hand by nature. To put it in very general terms, insurance is a better framework for diseases which can in principle be cured than it is for chronic conditions where future outlays are known with a great level of certainty; the latter type of disease tends to be difficult to handle in an insurance context.

People who have one disease may develop other diseases as time progresses, and having disease X may increase or decrease the risk of disease Y. People study such disease variability patterns, and have done so for years, but there’s still a lot of stuff we don’t know – here’s a recent post on these topics. Such patterns are interesting for multiple reasons. One major motivation for studying these things is that ‘different’ diseases may have common mechanisms, and the identification of these mechanisms may lead to new treatment options. A completely different motivation for studying these things relate rather to the kind of stuff covered in this post, where you instead wonder about economic aspects; for example, if the smoker stops smoking he may gain weight and eventually develop type 2 diabetes instead of developing some smoking-related condition. Is this outcome better or worse than the other? It’s important to keep in mind when evaluating changes in compensation schedules/insurance structures that diseases are not independent, and this is a problem regardless of whether you’re interested in total costs or ‘just’ direct outlays. Say you’re ‘only’ worried about outlays and you are trying to figure out if it is a good idea to deny coverage to smokers, and you know that ex-smokers are likely to gain weight and have an increased risk of type 2 diabetes. Then the relevant change in cost is not the money you save on smoking-related illness, it’s the cost change you arrive at when after you account for those savings also account for the increased cost of treating type 2 diabetes. Disease interdependencies are probably as complex as risk factor interdependencies – the two phenomena are to some extent representing the same basic phenomenon – so this makes true cost evaluation even harder than it already was. Not all relevant costs at the societal level are of course medical costs; if people live longer, and they rely partly on a pension scheme to which they are no longer contributing, that cost is also relevant.

If a group of people who live longer cost more than a group of people who do not live as long, and you need to cover the associated shortfall, then – as we concluded in the beginning – there are really only two ways to handle this: Make them pay more than the people who do not live as long, or make the people who do not live as long pay more to cover the shortfall. Another way to look at this is that in this situation you can either tax people ‘for not living long enough’, or you can tax people for ‘not dying at the appropriate time’. On the other hand (?), if a group of people who die early turns out to be the higher-cost group in the relevant comparison (perhaps because they have shorter working lives and so pay into the system for a shorter amount of time), then you can deal with this problem by… either taxing them for ‘not living long enough’ or by punishing the people who live long lives for ‘not dying at the appropriate time’. No, of course it doesn’t matter which group is high cost, the solution mechanism is the same in both cases – make one of the groups pay more. And every time you tweak things you change the incentives of various people, and implicit effects like these hide somewhere in the background.

March 31, 2017 Posted by | Cancer/oncology, Diabetes, Economics, health care, rambling nonsense | Leave a comment


All the quotes included in this post are from The Faber Book of Aphorisms, which I am currently reading.

i. “It is never any good dwelling on good-bys. It is not the being together that it prolongs, it is the parting.” (Elizabeth Bibesco)

ii. “Good manners are made up of petty sacrifices.” (Ralph Waldo Emerson)

iii. “One learns taciturnity best among people without it, and loquacity among the taciturn.” (Jean Paul Richter)

iv. “A man never reveals his character more vividly than when portraying the character of another.” (-ll-)

v. “That we seldom repent of talking too little and very often of talking too much is a … maxim that everybody knows and nobody practices.” (Jean de La Bruyère)

vi. “Never trust a man who speaks well of everybody.” (John Churton Collins)

vii. “People not used to the world … are unskillful enough to show what they have sense enough not to tell.” (Philip Dormer Stanhope, 4th Earl of Chesterfield)

viii. “To most men, experience is like the stern lights of a ship, which illumine only the track it has passed.” (Samuel Taylor Coleridge)

ix. “Those who know the least obey the best.” (George Farquhar)

x. “Monkeys are superior to men in this: when a monkey looks into a mirror, he sees a monkey.” (Malcolm de Chazal)

xi. “It can be shown that a mathematical web of some kind can be woven about any universe containing several objects. The fact that our universe lends itself to mathematical treatment is not a fact of any great philosophical significance.” (Bertrand Russell)

xii. “You can change your faith without changing gods, and vice versa.” (Stanisław Jerzy Lec)

xiii. “Religion is the masterpiece of the art of animal training, for it trains people as to how they shall think.” (Arthur Schopenhauer)

xiv. “The vanity of being known to be trusted with a secret is generally one of the chief motives to disclose it.” (Samuel Johnson)

xv. “No man is exempt from saying silly things; the mischief is to say them deliberately.” (Michel de Montaigne)

xvi. “Many promising reconciliations have broken down because, while both parties came prepared to forgive, neither party came prepared to be forgiven.” (Charles Williams)

xvii. “Ambition is pitiless. Any merit that it cannot use it finds despicable.” (Joseph Joubert)

xviii. “Experience is the name everyone gives to his mistakes.” (Oscar Wilde)

xix. “Nothing is enough to the man for whom enough is too little.” (Epicurus)

xx. “To measure up to all that is demanded of him, a man must overestimate his capacities.” (Johann Wolfgang von Goethe)

March 27, 2017 Posted by | Books, Quotes/aphorisms | Leave a comment

Promoting the unknown…

March 24, 2017 Posted by | Music | Leave a comment


Over the last couple of weeks I’ve been reading James Herriot’s books and yesterday I finished the last one in the series. The five books (or 8, if you’re British – see the wiki…) I read – I skipped the ‘dog stories’ publication on that list because that book is just a collection of stories included in the other books – contain almost 2500 pages (2479, according to the goodreads numbers provided in the context of the editions I’ve been reading), and they also contained quite a few unfamiliar/nice words and expressions, many of which are included below. If you’re curious about the Herriot books you can read my goodreads reviews of the books here, here (very short), here, and here (I didn’t review The Lord God Made Them All).

Eversionskeevy, censerknout, byreelectuary, trocar/trocarization, clogirascible, gilt, curvet, bullock, niggle, scapegrace, cur, pantile, raddle, scamper, skitter, odoriferous.

Dewlap, seton, muzzy, stirk, shillelagh, borborygmi, omentum, fettle, guddle, cruciate, peduncle/pedunculated, ecraseur, curlew, gabble, gable, festoon, cornada, lambent, lank.

Lope, billet, casement, scree, caliper, dale, stoup, puisne, tumefy, scamp, probang, famble, footling, colostrum, towsle/tousle, loquacious, dapper, cob, meconium, locum.

Mullion, roan, slat, dustman, carvery, abomasum, rostrum, zareba, flithackle, tympanites, pewter, opisthotonos, concertina, miliarylief, spay, otodectic.

March 24, 2017 Posted by | Books, language | Leave a comment


i. “Don’t be yourself. Be someone a little nicer.” (Mignon McLaughlin)

ii. “There are so many things that we wish we had done yesterday, so few that we feel like doing today.” (-ll-)

iii. “It’s the most unhappy people who most fear change.” (-ll-)

iv. “The essence of good manners consists in making it clear that one has no wish to hurt. When it is clearly necessary to hurt, it must be done in such a way as to make it evident that the necessity is felt to be regrettable.” (Bertrand Russell)

v. “Pretexts are not wanting when one wishes to use them.” (Carlo Goldoni)

vi. “In disputes upon moral or scientific points, ever let your aim be to come at truth, not to conquer your opponent: so you never shall be at a loss in losing the argument, and gaining a new discovery.” (James Burgh)

vii. “Be sure of the fact before you lose time in searching for a cause.” (-ll-)

viii. “The modest man is seldom the object of envy.” (-ll-)

ix. “Too much company is worse than none.” (-ll-)

x. “Men more frequently require to be reminded than informed.” (Samuel Johnson)

xi. “We are far more liable to catch the vices than the virtues of our associates.” (Denis Diderot)

xii. “We should measure our wealth according to the means we have of satisfying our desires.” (Antoine Francois Prevost d’Exiles)

xiii. “That virtue we appreciate is as much ours as another’s. We see so much only as we possess.” (Henry David Thoreau)

xiv. “Man soon finds what he wants to find. If he cannot find it otherwise, he creates it for his special enjoyment” (Alexander Bryan Johnson)

xv. “Whenever a man’s friends begin to compliment him about looking young, he may be sure that they think he is growing old.” (Washington Irving)

xvi. “It has been said that there is nothing more uncommon than common sense.” (Thomas Chalmers)

xvii. “If it had not been for the discontent of a few fellows who had not been satisfied with their conditions, you would still be living in caves. Intelligent discontent is the mainspring of civilization.” (Eugene Victor Debs)

xviii. “Often, the surest way to convey misinformation is to tell the strict truth.” (Mark Twain)

xviii. “He who lacks a single tael sees many bargains” (Ernest Bramah, Kai Lung’s Golden Hours)

xix. “Although there exist many thousand subjects for elegant conversation, there are persons who cannot meet a cripple without talking about feet.” (Ernest Bramah, The Wallet of Kai Lung)

xx. “When a lonely, penniless old woman dies people don’t rush up to you in the street to tell you.” (James Herriot, All Creatures Great and Small)

March 16, 2017 Posted by | Quotes/aphorisms | Leave a comment

Information complexity and applications

I have previously here on the blog posted multiple lectures in my ‘lecture-posts’, or I have combined a lecture with other stuff (e.g. links such as those in the previous ‘random stuff’ post). I think such approaches have made me less likely to post lectures on the blog (if I don’t post a lecture soon after I’ve watched it, my experience tells me that I not infrequently simply never get around to posting it), and combined with this issue is also the issue that I don’t really watch a lot of lectures these days. For these reasons I have decided to start posting single lecture posts here on the blog; when I start thinking about the time expenditure of people reading along here in a way this approach actually also seems justified – although it might take me as much time/work to watch and cover, say, 4 lectures as it would take me to read and cover 100 pages of a textbook, the time expenditure required by a reader of the blog would be very different in those two cases (you’ll usually be able to read a post that took me multiple hours to write in a short amount of time, whereas ‘the time advantage’ of the reader is close to negligible (maybe not completely; search costs are not completely irrelevant) in the case of lectures). By posting multiple lectures in the same post I probably decrease the expected value of the time readers spend watching the content I upload, which seems suboptimal.

Here’s the youtube description of the lecture, which was posted a few days ago on the IAS youtube account:

“Over the past two decades, information theory has reemerged within computational complexity theory as a mathematical tool for obtaining unconditional lower bounds in a number of models, including streaming algorithms, data structures, and communication complexity. Many of these applications can be systematized and extended via the study of information complexity – which treats information revealed or transmitted as the resource to be conserved. In this overview talk we will discuss the two-party information complexity and its properties – and the interactive analogues of classical source coding theorems. We will then discuss applications to exact communication complexity bounds, hardness amplification, and quantum communication complexity.”

He actually decided to skip the quantum communication complexity stuff because of the time constraint. I should note that the lecture was ‘easy enough’ for me to follow most of it, so it is not really that difficult, at least not if you know some basic information theory.

A few links to related stuff (you can take these links as indications of what sort of stuff the lecture is about/discusses, if you’re on the fence about whether or not to watch it):
Computational complexity theory.
Shannon entropy.
Shannon’s source coding theorem.
Communication complexity.
Communications protocol.
Information-based complexity.
Hash function.
From Information to Exact Communication (in the lecture he discusses some aspects covered in this paper).
Unique games conjecture (Two-prover proof systems).
A Counterexample to Strong Parallel Repetition (another paper mentioned/briefly discussed during the lecture).
Pinsker’s inequality.

An interesting aspect I once again noted during this lecture is the sort of loose linkage you sometimes observe between the topics of game theory/microeconomics and computer science. Of course the link is made explicit a few minutes later in the talk when he discusses the unique games conjecture to which I link above, but it’s perhaps worth noting that the link is on display even before that point is reached. Around 38 minutes into the lecture he mentions that one of the relevant proofs ‘involves such things as Lagrange multipliers and optimization’. I was far from surprised, as from a certain point of view the problem he discusses at that point is conceptually very similar to some problems encountered in auction theory, where Lagrange multipliers and optimization problems are frequently encountered… If you are too unfamiliar with that field to realize how the similar problem might appear in an auction theory context, what you have there are instead auction partipants who prefer not to reveal their true willingness to pay; and some auction designs actually work in a very similar manner as does the (pseudo-)protocol described in the lecture, and are thus used to reveal it (for some subset of participants at least)).

March 12, 2017 Posted by | Computer science, Game theory, Lectures, Papers | Leave a comment

Random stuff

It’s been a long time since I last posted one of these posts, so a great number of links of interest has accumulated in my bookmarks. I intended to include a large number of these in this post and this of course means that I surely won’t cover each specific link included in this post in anywhere near the amount of detail it deserves, but that can’t be helped.

i. Autism Spectrum Disorder Grown Up: A Chart Review of Adult Functioning.

“For those diagnosed with ASD in childhood, most will become adults with a significant degree of disability […] Seltzer et al […] concluded that, despite considerable heterogeneity in social outcomes, “few adults with autism live independently, marry, go to college, work in competitive jobs or develop a large network of friends”. However, the trend within individuals is for some functional improvement over time, as well as a decrease in autistic symptoms […]. Some authors suggest that a sub-group of 15–30% of adults with autism will show more positive outcomes […]. Howlin et al. (2004), and Cederlund et al. (2008) assigned global ratings of social functioning based on achieving independence, friendships/a steady relationship, and education and/or a job. These two papers described respectively 22% and 27% of groups of higher functioning (IQ above 70) ASD adults as attaining “Very Good” or “Good” outcomes.”

“[W]e evaluated the adult outcomes for 45 individuals diagnosed with ASD prior to age 18, and compared this with the functioning of 35 patients whose ASD was identified after 18 years. Concurrent mental illnesses were noted for both groups. […] Comparison of adult outcome within the group of subjects diagnosed with ASD prior to 18 years of age showed significantly poorer functioning for those with co-morbid Intellectual Disability, except in the domain of establishing intimate relationships [my emphasis. To make this point completely clear, one way to look at these results is that apparently in the domain of partner-search autistics diagnosed during childhood are doing so badly in general that being intellectually disabled on top of being autistic is apparently conferring no additional disadvantage]. Even in the normal IQ group, the mean total score, i.e. the sum of the 5 domains, was relatively low at 12.1 out of a possible 25. […] Those diagnosed as adults had achieved significantly more in the domains of education and independence […] Some authors have described a subgroup of 15–27% of adult ASD patients who attained more positive outcomes […]. Defining an arbitrary adaptive score of 20/25 as “Good” for our normal IQ patients, 8 of thirty four (25%) of those diagnosed as adults achieved this level. Only 5 of the thirty three (15%) diagnosed in childhood made the cutoff. (The cut off was consistent with a well, but not superlatively, functioning member of society […]). None of the Intellectually Disabled ASD subjects scored above 10. […] All three groups had a high rate of co-morbid psychiatric illnesses. Depression was particularly frequent in those diagnosed as adults, consistent with other reports […]. Anxiety disorders were also prevalent in the higher functioning participants, 25–27%. […] Most of the higher functioning ASD individuals, whether diagnosed before or after 18 years of age, were functioning well below the potential implied by their normal range intellect.”

Related papers: Social Outcomes in Mid- to Later Adulthood Among Individuals Diagnosed With Autism and Average Nonverbal IQ as Children, Adults With Autism Spectrum Disorders.

ii. Premature mortality in autism spectrum disorder. This is a Swedish matched case cohort study. Some observations from the paper:

“The aim of the current study was to analyse all-cause and cause-specific mortality in ASD using nationwide Swedish population-based registers. A further aim was to address the role of intellectual disability and gender as possible moderators of mortality and causes of death in ASD. […] Odds ratios (ORs) were calculated for a population-based cohort of ASD probands (n = 27 122, diagnosed between 1987 and 2009) compared with gender-, age- and county of residence-matched controls (n = 2 672 185). […] During the observed period, 24 358 (0.91%) individuals in the general population died, whereas the corresponding figure for individuals with ASD was 706 (2.60%; OR = 2.56; 95% CI 2.38–2.76). Cause-specific analyses showed elevated mortality in ASD for almost all analysed diagnostic categories. Mortality and patterns for cause-specific mortality were partly moderated by gender and general intellectual ability. […] Premature mortality was markedly increased in ASD owing to a multitude of medical conditions. […] Mortality was significantly elevated in both genders relative to the general population (males: OR = 2.87; females OR = 2.24)”.

“Individuals in the control group died at a mean age of 70.20 years (s.d. = 24.16, median = 80), whereas the corresponding figure for the entire ASD group was 53.87 years (s.d. = 24.78, median = 55), for low-functioning ASD 39.50 years (s.d. = 21.55, median = 40) and high-functioning ASD 58.39 years (s.d. = 24.01, median = 63) respectively. […] Significantly elevated mortality was noted among individuals with ASD in all analysed categories of specific causes of death except for infections […] ORs were highest in cases of mortality because of diseases of the nervous system (OR = 7.49) and because of suicide (OR = 7.55), in comparison with matched general population controls.”

iii. Adhesive capsulitis of shoulder. This one is related to a health scare I had a few months ago. A few quotes:

Adhesive capsulitis (also known as frozen shoulder) is a painful and disabling disorder of unclear cause in which the shoulder capsule, the connective tissue surrounding the glenohumeral joint of the shoulder, becomes inflamed and stiff, greatly restricting motion and causing chronic pain. Pain is usually constant, worse at night, and with cold weather. Certain movements or bumps can provoke episodes of tremendous pain and cramping. […] People who suffer from adhesive capsulitis usually experience severe pain and sleep deprivation for prolonged periods due to pain that gets worse when lying still and restricted movement/positions. The condition can lead to depression, problems in the neck and back, and severe weight loss due to long-term lack of deep sleep. People who suffer from adhesive capsulitis may have extreme difficulty concentrating, working, or performing daily life activities for extended periods of time.”

Some other related links below:

The prevalence of a diabetic condition and adhesive capsulitis of the shoulder.
“Adhesive capsulitis is characterized by a progressive and painful loss of shoulder motion of unknown etiology. Previous studies have found the prevalence of adhesive capsulitis to be slightly greater than 2% in the general population. However, the relationship between adhesive capsulitis and diabetes mellitus (DM) is well documented, with the incidence of adhesive capsulitis being two to four times higher in diabetics than in the general population. It affects about 20% of people with diabetes and has been described as the most disabling of the common musculoskeletal manifestations of diabetes.”

Adhesive Capsulitis (review article).
“Patients with type I diabetes have a 40% chance of developing a frozen shoulder in their lifetimes […] Dominant arm involvement has been shown to have a good prognosis; associated intrinsic pathology or insulin-dependent diabetes of more than 10 years are poor prognostic indicators.15 Three stages of adhesive capsulitis have been described, with each phase lasting for about 6 months. The first stage is the freezing stage in which there is an insidious onset of pain. At the end of this period, shoulder ROM [range of motion] becomes limited. The second stage is the frozen stage, in which there might be a reduction in pain; however, there is still restricted ROM. The third stage is the thawing stage, in which ROM improves, but can take between 12 and 42 months to do so. Most patients regain a full ROM; however, 10% to 15% of patients suffer from continued pain and limited ROM.”

Musculoskeletal Complications in Type 1 Diabetes.
“The development of periarticular thickening of skin on the hands and limited joint mobility (cheiroarthropathy) is associated with diabetes and can lead to significant disability. The objective of this study was to describe the prevalence of cheiroarthropathy in the well-characterized Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) cohort and examine associated risk factors […] This cross-sectional analysis was performed in 1,217 participants (95% of the active cohort) in EDIC years 18/19 after an average of 24 years of follow-up. Cheiroarthropathy — defined as the presence of any one of the following: adhesive capsulitis, carpal tunnel syndrome, flexor tenosynovitis, Dupuytren’s contracture, or a positive prayer sign [related link] — was assessed using a targeted medical history and standardized physical examination. […] Cheiroarthropathy was present in 66% of subjects […] Cheiroarthropathy is common in people with type 1 diabetes of long duration (∼30 years) and is related to longer duration and higher levels of glycemia. Clinicians should include cheiroarthropathy in their routine history and physical examination of patients with type 1 diabetes because it causes clinically significant functional disability.”

Musculoskeletal disorders in diabetes mellitus: an update.
“Diabetes mellitus (DM) is associated with several musculoskeletal disorders. […] The exact pathophysiology of most of these musculoskeletal disorders remains obscure. Connective tissue disorders, neuropathy, vasculopathy or combinations of these problems, may underlie the increased incidence of musculoskeletal disorders in DM. The development of musculoskeletal disorders is dependent on age and on the duration of DM; however, it has been difficult to show a direct correlation with the metabolic control of DM.”

Rheumatic Manifestations of Diabetes Mellitus.

Prevalence of symptoms and signs of shoulder problems in people with diabetes mellitus.

Musculoskeletal Disorders of the Hand and Shoulder in Patients with Diabetes.
“In addition to micro- and macroangiopathic complications, diabetes mellitus is also associated with several musculoskeletal disorders of the hand and shoulder that can be debilitating (1,2). Limited joint mobility, also termed diabetic hand syndrome or cheiropathy (3), is characterized by skin thickening over the dorsum of the hands and restricted mobility of multiple joints. While this syndrome is painless and usually not disabling (2,4), other musculoskeletal problems occur with increased frequency in diabetic patients, including Dupuytren’s disease [“Dupuytren’s disease […] may be observed in up to 42% of adults with diabetes mellitus, typically in patients with long-standing T1D” – link], carpal tunnel syndrome [“The prevalence of [carpal tunnel syndrome, CTS] in patients with diabetes has been estimated at 11–30 % […], and is dependent on the duration of diabetes. […] Type I DM patients have a high prevalence of CTS with increasing duration of disease, up to 85 % after 54 years of DM” – link], palmar flexor tenosynovitis or trigger finger [“The incidence of trigger finger [/stenosing tenosynovitis] is 7–20 % of patients with diabetes comparing to only about 1–2 % in nondiabetic patients” – link], and adhesive capsulitis of the shoulder (5–10). The association of adhesive capsulitis with pain, swelling, dystrophic skin, and vasomotor instability of the hand constitutes the “shoulder-hand syndrome,” a rare but potentially disabling manifestation of diabetes (1,2).”

“The prevalence of musculoskeletal disorders was greater in diabetic patients than in control patients (36% vs. 9%, P < 0.01). Adhesive capsulitis was present in 12% of the diabetic patients and none of the control patients (P < 0.01), Dupuytren’s disease in 16% of diabetic and 3% of control patients (P < 0.01), and flexor tenosynovitis in 12% of diabetic and 2% of control patients (P < 0.04), while carpal tunnel syndrome occurred in 12% of diabetic patients and 8% of control patients (P = 0.29). Musculoskeletal disorders were more common in patients with type 1 diabetes than in those with type 2 diabetes […]. Forty-three patients [out of 100] with type 1 diabetes had either hand or shoulder disorders (37 with hand disorders, 6 with adhesive capsulitis of the shoulder, and 10 with both syndromes), compared with 28 patients [again out of 100] with type 2 diabetes (24 with hand disorders, 4 with adhesive capsulitis of the shoulder, and 3 with both syndromes, P = 0.03).”

Association of Diabetes Mellitus With the Risk of Developing Adhesive Capsulitis of the Shoulder: A Longitudinal Population-Based Followup Study.
“A total of 78,827 subjects with at least 2 ambulatory care visits with a principal diagnosis of DM in 2001 were recruited for the DM group. The non-DM group comprised 236,481 age- and sex-matched randomly sampled subjects without DM. […] During a 3-year followup period, 946 subjects (1.20%) in the DM group and 2,254 subjects (0.95%) in the non-DM group developed ACS. The crude HR of developing ACS for the DM group compared to the non-DM group was 1.333 […] the association between DM and ACS may be explained at least in part by a DM-related chronic inflammatory process with increased growth factor expression, which in turn leads to joint synovitis and subsequent capsular fibrosis.”

It is important to note when interpreting the results of the above paper that these results are based on Taiwanese population-level data, and type 1 diabetes – which is obviously the high-risk diabetes subgroup in this particular context – is rare in East Asian populations (as observed in Sperling et al., “A child in Helsinki, Finland is almost 400 times more likely to develop diabetes than a child in Sichuan, China”. Taiwanese incidence of type 1 DM in children is estimated at ~5 in 100.000).

iv. Parents who let diabetic son starve to death found guilty of first-degree murder. It’s been a while since I last saw one of these ‘boost-your-faith-in-humanity’-cases, but they in my impression do pop up every now and then. I should probably keep at hand one of these articles in case my parents ever express worry to me that they weren’t good parents; they could have done a lot worse…

v. Freedom of medicine. One quote from the conclusion of Cochran’s post:

“[I]t is surely possible to materially improve the efficacy of drug development, of medical research as a whole. We’re doing better than we did 500 years ago – although probably worse than we did 50 years ago. But I would approach it by learning as much as possible about medical history, demographics, epidemiology, evolutionary medicine, theory of senescence, genetics, etc. Read Koch, not Hayek. There is no royal road to medical progress.”

I agree, and I was considering including some related comments and observations about health economics in this post – however I ultimately decided against doing that in part because the post was growing unwieldy; I might include those observations in another post later on. Here’s another somewhat older Westhunt post I at some point decided to bookmark – I in particular like the following neat quote from the comments, which expresses a view I have of course expressed myself in the past here on this blog:

“When you think about it, falsehoods, stupid crap, make the best group identifiers, because anyone might agree with you when you’re obviously right. Signing up to clear nonsense is a better test of group loyalty. A true friend is with you when you’re wrong. Ideally, not just wrong, but barking mad, rolling around in your own vomit wrong.”

vi. Economic Costs of Diabetes in the U.S. in 2012.

“Approximately 59% of all health care expenditures attributed to diabetes are for health resources used by the population aged 65 years and older, much of which is borne by the Medicare program […]. The population 45–64 years of age incurs 33% of diabetes-attributed costs, with the remaining 8% incurred by the population under 45 years of age. The annual attributed health care cost per person with diabetes […] increases with age, primarily as a result of increased use of hospital inpatient and nursing facility resources, physician office visits, and prescription medications. Dividing the total attributed health care expenditures by the number of people with diabetes, we estimate the average annual excess expenditures for the population aged under 45 years, 45–64 years, and 65 years and above, respectively, at $4,394, $5,611, and $11,825.”

“Our logistic regression analysis with NHIS data suggests that diabetes is associated with a 2.4 percentage point increase in the likelihood of leaving the workforce for disability. This equates to approximately 541,000 working-age adults leaving the workforce prematurely and 130 million lost workdays in 2012. For the population that leaves the workforce early because of diabetes-associated disability, we estimate that their average daily earnings would have been $166 per person (with the amount varying by demographic). Presenteeism accounted for 30% of the indirect cost of diabetes. The estimate of a 6.6% annual decline in productivity attributed to diabetes (in excess of the estimated decline in the absence of diabetes) equates to 113 million lost workdays per year.”

vii. Total red meat intake of ≥0.5 servings/d does not negatively influence cardiovascular disease risk factors: a systemically searched meta-analysis of randomized controlled trials.

viii. Effect of longer term modest salt reduction on blood pressure: Cochrane systematic review and meta-analysis of randomised trials. Did I blog this paper at some point in the past? I could not find any coverage of it on the blog when I searched for it so I decided to include it here, even if I have a nagging suspicion I may have talked about these findings before. What did they find? The short version is this:

“A modest reduction in salt intake for four or more weeks causes significant and, from a population viewpoint, important falls in blood pressure in both hypertensive and normotensive individuals, irrespective of sex and ethnic group. Salt reduction is associated with a small physiological increase in plasma renin activity, aldosterone, and noradrenaline and no significant change in lipid concentrations. These results support a reduction in population salt intake, which will lower population blood pressure and thereby reduce cardiovascular disease.”

ix. Some wikipedia links:

Heroic Age of Antarctic Exploration (featured).

Wien’s displacement law.

Kuiper belt (featured).

Treason (one quote worth including here: “Currently, the consensus among major Islamic schools is that apostasy (leaving Islam) is considered treason and that the penalty is death; this is supported not in the Quran but in the Hadith.[42][43][44][45][46][47]“).

Lymphatic filariasis.

File:World map of countries by number of cigarettes smoked per adult per year.

Australian gold rushes.

Savant syndrome (“It is estimated that 10% of those with autism have some form of savant abilities”). A small sidenote of interest to Danish readers: The Danish Broadcasting Corporation recently featured a series about autistics with ‘special abilities’ – the show was called ‘The hidden talents’ (De skjulte talenter), and after multiple people had nagged me to watch it I ended up deciding to do so. Most of the people in that show presumably had some degree of ‘savantism’ combined with autism at the milder end of the spectrum, i.e. Asperger’s. I was somewhat conflicted about what to think about the show and did consider blogging it in detail (in Danish?), but I decided against it. However I do want to add here to Danish readers reading along who’ve seen the show that they would do well to repeatedly keep in mind that a) the great majority of autistics do not have abilities like these, b) many autistics with abilities like these presumably do quite poorly, and c) that many autistics have even greater social impairments than do people like e.g. (the very likeable, I have to add…) Louise Wille from the show).

Quark–gluon plasma.

Simo Häyhä.

Chernobyl liquidators.

Black Death (“Over 60% of Norway’s population died in 1348–1350”).

Renault FT (“among the most revolutionary and influential tank designs in history”).

Weierstrass function (“an example of a pathological real-valued function on the real line. The function has the property of being continuous everywhere but differentiable nowhere”).

W Ursae Majoris variable.

Void coefficient. (“a number that can be used to estimate how much the reactivity of a nuclear reactor changes as voids (typically steam bubbles) form in the reactor moderator or coolant. […] Reactivity is directly related to the tendency of the reactor core to change power level: if reactivity is positive, the core power tends to increase; if it is negative, the core power tends to decrease; if it is zero, the core power tends to remain stable. […] A positive void coefficient means that the reactivity increases as the void content inside the reactor increases due to increased boiling or loss of coolant; for example, if the coolant acts as a neutron absorber. If the void coefficient is large enough and control systems do not respond quickly enough, this can form a positive feedback loop which can quickly boil all the coolant in the reactor. This happened in the RBMK reactor that was destroyed in the Chernobyl disaster.”).

Gregor MacGregor (featured) (“a Scottish soldier, adventurer, and confidence trickster […] MacGregor’s Poyais scheme has been called one of the most brazen confidence tricks in history.”).


Irish Civil War.

March 10, 2017 Posted by | Astronomy, autism, Cardiology, Diabetes, Economics, Epidemiology, History, Infectious disease, Mathematics, Medicine, Papers, Physics, Psychology, Random stuff, Wikipedia | Leave a comment


i. “Fraud and falsehood only dread examination. Truth invites it.” (Thomas Cooper)

ii. “However well equipped our language, it can never be forearmed against all possible cases that may arise and call for description: fact is richer than diction.” (J. L. Austin)

iii. “There is no loneliness like the loneliness of crowds, especially to those who are unaccustomed to them.” (H. Rider Haggard)

iv. “All men are moral. Only their neighbors are not.” (John Steinbeck)

v. “The unfortunate thing is that, because wishes sometimes come true, the agony of hoping is perpetuated.” (Marguerite Cleenewerck de Crayencour)

vi. “All cruel people describe themselves as paragons of frankness.” (Tennessee Williams)

vii. “If you do not have the capacity for happiness with a little money, great wealth will not bring it to you.” (William Feather)

viii. “Anyone who can think clearly can write clearly. But neither is easy.” (-ll-)

ix. “No one’s reputation is quite what he himself perceives it ought to be.” (Christopher Vokes)

x. “[T]he question is not how to avoid procrastination, but how to procrastinate well. There are three variants of procrastination, depending on what you do instead of working on something: you could work on (a) nothing, (b) something less important, or (c) something more important. That last type, I’d argue, is good procrastination.” (Paul Graham)

xi. “At every period of history, people have believed things that were just ridiculous, and believed them so strongly that you risked ostracism or even violence by saying otherwise. If our own time were any different, that would be remarkable. As far as I can tell it isn’t.” (-ll-)

xii. “There can be no doubt that the knowledge of logic is of considerable practical importance for everyone who desires to think and infer correctly.” (Alfred Tarski)

xiii. “Logic and truth are two very different things, but they often look the same to the mind that’s performing the logic.” (Theodore Sturgeon)

xiv.”I don’t like it; I can’t approve of it; I have always thought it most regrettable that earnest and ethical Thinkers like ourselves should go scuttling through space in this undignified manner. Is it seemly that I, at my age, should be hurled with my books of reference, and bed-clothes, and hot-water bottle, across the sky at the unthinkable rate of nineteen miles a second? As I say, I don’t at all like it.” (Logan Pearsall Smith, All Trivia).

xv. “That we should practice what we preach is generally admitted; but anyone who preaches what he and his hearers practise must incur the gravest moral disapprobation.” (-ll-)

xvi. “Our names are labels, plainly printed on the bottled essense of our past behaviour.” (-ll-)

xvii. “It’s an odd thing about this Universe that though we all disagree with each other, we are all of us always in the right.” (-ll-)

xviii. “Those who say everything is pleasant and everyone delightful, come to the awful fate of believing what they say.” (-ll-)

xix. “He who goes against the fashion is himself its slave.” (-ll-)

xx. “When I read in the Times about India and all its problems and populations; when I look at the letters in large type of important personages, and find myself face to face with the Questions, Movements, and great Activities of the Age, ‘Where do I come in?’ I ask uneasily.
Then in the great Times-reflected world I find the corner where I play my humble but necessary part. For I am one of the unpraised, unrewarded millions without whom Statistics would be a bankrupt science. It is we who are born, who marry, who die, in constant ratios; who regularly lose so many umbrellas, post just so many unaddressed letters every year. And there are enthusiasts among us, Heroes who, without the least thought of their own convenience, allow omnibuses to run over them, or throw themselves, great-heartedly, month by month, in fixed numbers, from London bridges.” (-ll-)

March 9, 2017 Posted by | Books, Quotes/aphorisms | Leave a comment

Biodemography of aging (II)

In my first post about the book I included a few general remarks about the book and what it’s about. In this post I’ll continue my coverage of the book, starting with a few quotes from and observations related to the content in chapter 4 (‘Evidence for Dependence Among Diseases‘).

“To compare the effects of public health policies on a population’s characteristics, researchers commonly estimate potential gains in life expectancy that would result from eradication or reduction of selected causes of death. For example, Keyfitz (1977) estimated that eradication of cancer would result in 2.265 years of increase in male life expectancy at birth (or by 3 % compared to its 1964 level). Lemaire (2005) found that the potential gain in the U.S. life expectancy from cancer eradication would not exceed 3 years for both genders. Conti et al. (1999) calculated that the potential gain in life expectancy from cancer eradication in Italy would be 3.84 years for males and 2.77 years for females. […] All these calculations assumed independence between cancer and other causes of death. […] for today’s populations in developed countries, where deaths from chronic non-communicable diseases are in the lead, this assumption might no longer be valid. An important feature of such chronic diseases is that they often develop in clusters manifesting positive correlations with each other. The conventional view is that, in a case of such dependence, the effect of cancer eradication on life expectancy would be even smaller.”

I think the great majority of people you asked would have assumed that the beneficial effect of hypothetical cancer eradication in humans on human life expectancy would be much larger than this, but that’s just an impression. I’ve seen estimates like these before, so I was not surprised – but I think many people would be if they knew this. A very large number of people die as a result of developing cancer today, but the truth of the matter is that if they hadn’t died from cancer they’d have died anyway, and on average probably not really all that much later. I linked to Richard Alexander’s comments on this topic in my last post about the book, and again his observations apply so I thought I might as well add the relevant quote from the book here:

“In the course of working against senescence, selection will tend to remove, one by one, the most frequent sources of mortality as a result of senescence. Whenever a single cause of mortality, such as a particular malfunction of any vital organ, becomes the predominant cause of mortality, then selection will more effectively reduce the significance of that particular defect (meaning those who lack it will outreproduce) until some other achieves greater relative significance. […] the result will be that all organs and systems will tend to deteriorate together. […] The point is that as we age, and as senescence proceeds, large numbers of potential sources of mortality tend to lurk ever more malevolently just “below the surface,”so that, unfortunately, the odds are very high against any dramatic lengthening of the maximum human lifetime through technology.”

Remove one cause of death and there are plenty of others standing in line behind it. We already knew that; two hundred years ago one out of every four deaths in England was the result of tuberculosis, but developing treatments for tuberculosis and other infectious diseases did not mean that English people stopped dying; these days they just die from cardiovascular disease and cancer instead. Do note in the context of that quote that Alexander is talking about the maximum human lifetime, not average life expectancy; again, we know and have known for a long time that human technology can have a dramatic effect on the latter variable. Of course a shift in one distribution will be likely to have spill-over effects on the other (if more people are alive at the age of 70, the potential group of people also living on to reach e.g. 100 years is higher, even if the mortality rate for the 70-100 year old group did not change) the point is just that these effects are secondary effects and are likely to be marginal at best.

Anyway, some more stuff from the chapter. Just like the previous chapter in the book did, this one also includes analyses of very large data sets:

The Multiple Cause of Death (MCD) data files contain information about underlying and secondary causes of death in the U.S. during 1968–2010. In total, they include more than 65 million individual death certificate records. […] we used data for the period 1979–2004.”

There’s some formal modelling stuff in the chapter which I won’t go into in detail here, this is the chapter in which I encountered the comment about ‘the multivariate lognormal frailty model’ I included in my first post about the book. One of the things the chapter looks at are the joint frequencies of deaths from cancer and other fatal diseases; it turns out that there are multiple diseases that are negatively related with cancer as a cause of death when you look at the population-level data mentioned above. The chapter goes into some of the biological mechanisms which may help explain why these associations look the way they do, and I’ll quote a little from that part of the coverage. A key idea here is (as always..?) that there are tradeoffs at play; some genetic variants may help protect you against e.g. cancer, but at the same time increase the risk of other diseases for the same reason that they protect you against cancer. In the context of the relationship between cancer deaths and deaths from other diseases they note in the conclusion that: “One potential biological mechanism underlying the negative correlation among cancer and other diseases could be related to the differential role of apoptosis in the development of these diseases.” The chapter covers that stuff in significantly more detail, and I decided to add some observations from the chapter on these topics below:

“Studying the role of the p53 gene in the connection between cancer and cellular aging, Campisi (2002, 2003) suggested that longevity may depend on a balance between tumor suppression and tissue renewal mechanisms. […] Although the mechanism by which p53 regulates lifespan remains to be determined, […] findings highlight the possibility that careful manipulation of p53 activity during adult life may result in beneficial effects on healthy lifespan. Other tumor suppressor genes are also involved in regulation of longevity. […] In humans, Dumont et al. (2003) demonstrated that a replacement of arginine (Arg) by proline (Pro) at position 72 of human p53 decreases its ability to initiate apoptosis, suggesting that these variants may differently affect longevity and vulnerability to cancer. Van Heemst et al. (2005) showed that individuals with the Pro/Pro genotype of p53 corresponding to reduced apoptosis in cells had significantly increased overall survival (by 41%) despite a more than twofold increased proportion of cancer deaths at ages 85+, together with a decreased proportion of deaths from senescence related causes such as COPD, fractures, renal failure, dementia, and senility. It was suggested that human p53 may protect against cancer but at a cost of longevity. […] Other biological factors may also play opposing roles in cancer and aging and thus contribute to respective trade-offs […]. E.g., higher levels of IGF-1 [have been] linked to both cancer and attenuation of phenotypes of physical senescence, such as frailty, sarcopenia, muscle atrophy, and heart failure, as well as to better muscle regeneration”.

“The connection between cancer and longevity may potentially be mediated by trade-offs between cancer and other diseases which do not necessarily involve any basic mechanism of aging per se. In humans, it could result, for example, from trade-offs between vulnerabilities to cancer and AD, or to cancer and CVD […] There may be several biological mechanisms underlying the negative correlation among cancer and these diseases. One can be related to the differential role of apoptosis in their development. For instance, in stroke, the number of dying neurons following brain ischemia (and thus probability of paralysis or death) may be less in the case of a downregulated apoptosis. As for cancer, the downregulated apoptosis may, conversely, mean a higher risk of the disease because more cells may survive damage associated with malignant transformation. […] Also, the role of the apoptosis may be different or even opposite in the development of cancer and Alzheimer’s disease (AD). Indeed, suppressed apoptosis is a hallmark of cancer, while increased apoptosis is a typical feature of AD […]. If so, then chronically upregulated apoptosis (e.g., due to a genetic polymorphism) may potentially be protective against cancer, but be deleterious in relation to AD. […] Increased longevity can be associated not only with increased but also with decreased chances of cancer. […] The most popular to-date “anti-aging” intervention, caloric restriction, often results in increased maximal life span along with reduced tumor incidence in laboratory rodents […] Because the rate of apoptosis was significantly and consistently higher in food restricted mice regardless of age, James et al. (1998) suggested that caloric restriction may have a cancer-protective effect primarily due to the upregulated apoptosis in these mice.”

Below I’ll discuss content covered in chapter 5, which deals with ‘Factors That May Increase Vulnerability to Cancer and Longevity in Modern Human Populations’. I’ll start out with a few quotes:

“Currently, the overall cancer incidence rate (age-adjusted) in the less developed world is roughly half that seen in the more developed world […] For countries with similar levels of economic development but different climate and ethnic characteristics […], the cancer rate patterns look much more similar than for the countries that share the same geographic location, climate, and ethnic distribution, but differ in the level of economic development […]. This suggests that different countries may share common factors linked to economic prosperity that could be primarily responsible for the modern increases in overall cancer risk. […] Population aging (increases in the proportion of older people) may […] partly explain the rise in the global cancer burden […]; however, it cannot explain increases in age-specific cancer incidence rates over time […]. Improved diagnostics and elevated exposures to carcinogens may explain increases in rates for selected cancer sites, but they cannot fully explain the increase in the overall cancer risk, nor incidence rate trends for most individual cancers (Jemal et al. 2008, 2013).”

“[W]e propose that the association between the overall cancer risk and the economic progress and spread of the Western lifestyle could in part be explained by the higher proportion of individuals more susceptible to cancer in the populations of developed countries, and discuss several mechanisms of such an increase in the proportion of the vulnerable. […] mechanisms include but are not limited to: (i) Improved survival of frail individuals. […] (ii) Avoiding or reducing traditional exposures. Excessive disinfection and hygiene typical of the developed world can diminish exposure to some factors that were abundant in the past […] Insufficiently or improperly trained immune systems may be less capable of resisting cancer. (iii) Burden of novel exposures. Some new medicines, cleaning agents, foods, etc., that are not carcinogenic themselves may still affect the natural ways of processing carcinogens in the body, and through this increase a person’s susceptibility to established carcinogens. [If this one sounds implausible to you, I’ll remind you that drug metabolism is complicatedUS] […] (iv) Some of the factors linked to economic prosperity and the Western lifestyle (e.g., delayed childbirth and food enriched with growth factors) may antagonistically influence aging and cancer risk.”

They provide detailed coverage of all of these mechanisms in the chapter, below I have included a few select observations from that part of the coverage.

“There was a dramatic decline in infant and childhood mortality in developed countries during the last century. For example, the infant mortality rate in the United States was about 6 % of live births in 1935, 3 % in 1950, 1.3 % in 1980, and 0.6 % in 2010. That is, it declined tenfold over the course of 75 years […] Because almost all children (including those with immunity deficiencies) survive, the proportion of the children who are inherently more vulnerable could be higher in the more developed countries. This is consistent with a typically higher proportion of children with chronic inflammatory immune disorders such as asthma and allergy in the populations of developed countries compared to less developed ones […] Over-reduction of such traditional exposures may result in an insufficiently/improperly trained immune system early in life, which could make it less able to resist diseases, including cancer later in life […] There is accumulating evidence of the important role of these effects in cancer risk. […] A number of studies have connected excessive disinfection and lack of antigenic stimulation (especially in childhood) of the immune system in Westernized communities with increased risks of both chronic inflammatory diseases and cancer […] The IARC data on migrants to Israel […] allow for comparison of the age trajectories of cancer incidence rates between adult Jews who live in Israel but were born in other countries […] [These data] show that Jews born in less developed regions (Africa and Asia) have overall lower cancer risk than those born in the more developed regions (Europe and America).  The discrepancy is unlikely to be due to differences in cancer diagnostics because at the moment of diagnosis all these people were citizens of the same country with the same standard of medical care. These results suggest that surviving childhood and growing up in a less developed country with diverse environmental exposures might help form resistance to cancer that lasts even after moving to a high risk country.”

I won’t go much into the ‘burden of novel exposures’ part, but I should note that exposures that may be relevant include factors like paracetamol use and antibiotics for treatment of H. pylori. Paracetamol is not considered carcinogenic by the IARC, but we know from animal studies that if you give rats paratamol and then expose them to an established carcinogen (with the straightforward name N-nitrosoethyl-N-hydroxyethylamine), the number of rats developing kidney cancer goes up. In the context of H. pylori, we know that these things may cause stomach cancer, but when you treat rats with metronidazol (which is used to treat H. pylori) and expose them to an established carcinogen, they’re more likely to develop colon cancer. The link between colon cancer and antibiotics use has been noted in other contexts as well; decreased microbial diversity after antibiotics use may lead to suppression of the bifidobacteria and promotion of E. coli in the colon, the metabolic products of which may lead to increased cancer risk. Over time an increase in colon cancer risk and a decrease in stomach cancer risk has been observed in developed societies, but aside from changes in diet another factor which may play a role is population-wide exposure to antibiotics. Colon and stomach cancers are incidentally not the only ones of interest in this particular context; it has also been found that exposure to chloramphenicol, a broad-spectrum antibiotic used since the 40es, increases the risk of lymphoma in mice when the mice are exposed to a known carcinogen, despite the drug itself again not being clearly carcinogenic on its own.

Many new exposures aside from antibiotics are of course relevant. Two other drug-related ones that might be worth mentioning are hormone replacement therapy and contraceptives. HRT is not as commonly used today as it was in the past, but to give some idea of the scope here, half of all women in the US aged 50-65 are estimated to have been on HRT at the peak of its use, around the turn of the millennium, and HRT is assumed to be partly responsible for the higher incidence of hormone-related cancers observed in female populations living in developed countries. It’s of some note that the use of HRT dropped dramatically shortly after this peak (from 61 million prescriptions in 2001 to 21 million in 2004), and that the incidence of estrogen-receptor positive cancers subsequently dropped. As for oral contraceptives, these have been in use since the 1960s, and combined hormonal contraceptives are known to increase the risk of liver- and breast cancer, while seemingly also having a protective effect against endometrial cancer and ovarian cancer. The authors speculate that some of the cancer incidence changes observed in the US during the latter half of the last century, with a decline in female endometrial and ovarian cancer combined with an increase in breast- and liver cancer, could in part be related to widespread use of these drugs. An estimated 10% of all women of reproductive age alive in the world, and 16% of those living in the US, are estimated to be using combined hormonal contraceptives. In the context of the protective effect of the drugs, it should perhaps be noted that endometrial cancer in particular is strongly linked to obesity so if you are not overweight you are relatively low-risk.

Many ‘exposures’ in a cancer context are not drug-related. For example women in Western societies tend to go into menopause at a higher age, and higher age of menopause has been associated with hormone-related cancers; but again the picture is not clear in terms of how the variable affects longevity, considering that later menopause has also been linked to increased longevity in several large studies. In the studies the women did have higher mortality from the hormone-related cancers, but on the other hand they were less likely to die from some of the other causes, such as pneumonia, influenza, and falls. Age of childbirth is also a variable where there are significant differences between developed countries and developing countries, and this variable may also be relevant to cancer incidence as it has been linked to breast cancer and melanoma; in one study women who first gave birth after the age of 35 had a 40% increased risk of breast cancer compared to mothers who gave birth before the age of 20 (good luck ‘controlling for everything’ in a context like that, but…), and in a meta-analysis the relative risk for melanoma was 1.47 for women in the oldest age group having given birth, compared to the youngest (again, good luck controlling for everything, but at least it’s not just one study). Lest you think this literature only deals with women, it’s also been found that parental age seems to be linked to cancers in the offspring (higher parental age -> higher cancer risk in the offspring), though the effect sizes are not mentioned in the coverage.

Here’s what they conclude at the end of the chapter:

“Some of the factors associated with economic prosperity and a Western lifestyle may influence both aging and vulnerability to cancer, sometimes oppositely. Current evidence supports a possibility of trade-offs between cancer and aging-related phenotypes […], which could be influenced by delayed reproduction and exposures to growth factors […]. The latter may be particularly beneficial at very old age. This is because the higher levels of growth factors may attenuate some phenotypes of physical senescence, such as decline in regenerative and healing ability, sarcopenia, frailty, elderly fractures and heart failure due to muscles athrophy. They may also increase the body’s vulnerability to cancer, e.g., through growth promoting and anti-apoptotic effects […]. The increase in vulnerability to cancer due to growth factors can be compatible with extreme longevity because cancer is a major contributor to mortality mainly before age 85, while senescence-related causes (such as physical frailty) become major contributors to mortality at oldest old ages (85+). In this situation, the impact of growth factors on vulnerability to death could be more deleterious in middle-to-old life (~before 85) and more beneficial at older ages (85+).

The complex relationships between aging, cancer, and longevity are challenging. This complexity warns against simplified approaches to extending longevity without taking into account the possible trade-offs between phenotypes of physical aging and various health disorders, as well as the differential impacts of such tradeoffs on mortality risks at different ages (e.g., Ukraintseva and Yashin 2003a; Yashin et al. 2009; Ukraintseva et al. 2010, 2016).”

March 7, 2017 Posted by | Books, Cancer/oncology, Epidemiology, Genetics, Immunology, Medicine, Pharmacology | Leave a comment


The great majority of the words below are from books I’ve recently read, I’ve almost not spent any time on since my last post of this kind; the guys add new words much too slowly, and most of the words they’ve recently added were not in my opinion all that interesting.

GelidCicatrization. Caudal. Stanchion. Saurian. Griddle. Branks. Purlieu. Arras. Slicker. Insipidity. Sedulously. Splay. Traipse. Gaff. Costive. Depauperate. Quaver. Homiletic.

Anemometer. Flagitious. CarboyMatutinalCognizance. Crispation. Doughty. Crepuscular. Giblets. Venery. Collier. Charnel. Dirge. Natter. Lintel. Disquisition.

Papuliferous. Vespertine. Lusciousness. Damask. Vaunt. Cicatrix. Integument. Heresiarch. Traducement. Apotheosis. Sardanapalian. Vocable. Desiderium. Leucocholy. Compathy.

Callosity. Skosh. Hellacious. Jouncy. Scilicet. Benignancy. TenebrificIpseityHoydenish. Quean. Handsel. Piton. Belvedere. Yenta. Officinal. Sanative. Umbra. Abaxial. Idiographic.

March 6, 2017 Posted by | language | 4 Comments

Economic Analysis in Healthcare (II)

This is my second and last post about the book, which will include some quotes from the second half of the book, as well as some comments.

“Different countries have adopted very different health care financing systems. In fact, it is arguable that the arrangements for financing of health care are more variable between different countries than the financing of any other good or service. […] The mechanisms adopted to deal with moral hazard are similar in all systems, whilst the mechanisms adopted to deal with adverse selection and incomplete coverage are very different. Compulsory insurance is used by social insurance and taxation [schemes] to combat adverse selection and incomplete coverage. Private insurance relies instead on experience rating to address adverse selection and a mix of retrospective reimbursement and selective contracting and vertical integration to deal with incomplete coverage.”

I have mentioned this before here on the blog (and elsewhere), but it is worth reiterating because you seem to sometimes encounter people who do not know this; there are some problems you’ll have to face when you’re dealing with insurance markets which will be there regardless of which entity is in charge of the insurance scheme. It doesn’t matter if your insurance system is government based or if the government is not involved in the insurance scheme at all, moral hazard will be there either way as a potential problem and you’re going to have to deal with that somehow. In econ 101 you tend to learn that ‘markets are great’, but this is one of those problems which are not going to go away by privatization.

On top of common problems faced by all insurers/insurance systems, different types of -systems will also tend to face a different mix of potential problems, some of which are likely to merit special attention in the specific setting in question. Some problems tend to be much more common in some specific settings than they are in others, which means that to some extent when you’re deciding on what might be ‘the ‘best’ institutional setup’, part of what you’re deciding on is which problem you are most concerned about addressing. In an evaluation context it should be pointed out in that context that the fact that most systems are mixes of different systems rather than ‘pure systems’, which they are, means that evaluation problems tend to be harder than they might otherwise have been. To add to this complexity as noted above the ways insurers deal with the same problem may not necessarily be the same in different institutional setups, which is worth having in mind when performance is evaluated (i.e., the fact that country A has included in the insurance system a feature X intending to address problem Q does not mean that country B, which has not included X in the system, does not attempt to address problem Q; B may just be using feature Y instead of feature X to do so).

Chapter 7 of the book deals with Equity in health care, and although I don’t want to cover that chapter in any detail a few observations from the text I did find worth including in this post:

“In the 1930s, only 43% of the [UK] population were covered by the national insurance scheme, mainly men in manual and low-paid occupations, and covered only for GP services. Around 21 million people were not covered by any health insurance, and faced potentially catastrophic expenditure should they become ill.”

“The literature on equity in the finance of health care has focused largely on the extent to which health care is financed according to ability to pay, and in particular on whether people with different levels of income make […] different payments, which is a vertical equity concern. Much less attention has been paid to horizontal equity, which considers the extent to which people with the same income make the same payments. […] There is horizontal inequity if people with the same ability to pay for health care, for example the same income, pay different amounts for it. […] tax-based payments and social health insurance payments tend to have less horizontal inequity than private health insurance payments and direct out-of-pocket payments. […] there are many concepts of equity that could be pursued; these are limited only by our capacity to think about the different ways in which resources could be allocated. It is unsurprising therefore that so many concepts of equity are discussed in the literature.”

Chapter 8 is about ‘Health care labour markets’. Again I won’t cover the chapter in much detail – people interested in such topics might like to have a look at this paper, which I concluded from a brief skim looks like it covers a few of the topics also discussed in the chapter – but I did want to include a few data:

“[S]alaries and wages paid to health care workers account for a substantial component of total health expenditure: the average country devotes over 40% of its government-funded health expenditure to paying its health workforce […], though there are regional variations [from ~30% in Africa to ~50% in the US and the Middle East – the data source is WHO, and the numbers are from 2006]. […] The WHO estimates there are around 59 million paid health workers worldwide […], around nine workers for every 1 000 population, with around two-thirds of the total providing health care and one third working in a non-clinical capacity.”

The last few chapters of the book cover mostly topics I have dealt with before, in more detail – for example are most topics covered here which are also covered in Gray et al. covered in much more detail in the latter book, which is natural as this text is mostly an introductory undergraduate text whereas the Gray et al. text is not (the latter book was based on material taught in a course called ‘Advanced Methods of Cost-Effectiveness Analysis’) – or topics in which I’m not actually all that interested (e.g. things like ‘extra-welfarism‘). Below I have added some quotes from the remaining chapters. I apologize in advance for repeating myself, given the fact that I probably covered a lot of this stuff back when I covered Gray et al., but on the other hand I read that book a while ago anyway:

“Simply providing information on costs and benefits is in itself not evaluative. Rather, in economic evaluation this information is structured in such a way as to enable alternative uses of resources to be judged. There are many criteria that might be used for such judgements. […] The criteria that are the focus of economic analysis are efficiency and equity […] in practice efficiency is dealt with far more often and with greater attention to precise numerical estimates. […] In publicly provided health programmes, market forces might be weak or there might be none at all. Economic evaluation is largely concerned with measuring efficiency in areas where there is public involvement and there are no markets to generate the kind of information – for example, prices and profits – that enable us to judge this. […] The question of how costs and benefits are to be measured and weighed against each other is obviously a fundamental issue, and indeed forms the main body of work on the topic. The answers to this question are often pragmatic, but they also have very strong guides from theory.”

“[M]any support economic evaluation as a useful technique even where it falls short of being a full cost–benefit analysis [‘CBA’ – US], as it provides at least some useful information. A partial cost–benefit analysis usually means that some aspects of cost or benefit have been identified but not valued, and the usefulness of the information depends on whether we believe that if the missing elements were to be valued they would alter the balance of costs and benefits. […] A special case of a partial economic evaluation is where costs are valued but benefits are not. […] This kind of partial efficiency is dealt with by a different type of economic evaluation known as cost-effectiveness analysis (CEA). […] One rationale for CEA is that whilst costs are usually measured in terms of money, it may be much more difficult to measure benefits that way. […] Cost-effectiveness analysis tries to identify where more benefit can be produced at the same cost or a lower cost can be achieved for the same benefit. […] there are many cases where we may wish to compare alternatives in which neither benefits nor costs are held constant. In this case, a cost-effectiveness ratio (CER) – the cost per unit of output or effect – is calculated to compare the alternatives, with the implication that the lower the CER the better. […] CBA seeks to answer whether or not a particular output is worth the cost. CEA seeks to answer the question of which among two or more alternatives provides the most output for a given cost, or the lowest cost for a given output. CBA therefore asks whether or not we should do things, while CEA asks what is the best way to do things that are worth doing.”

“The major preoccupation of economic evaluation in health care has been measurement of costs and benefits – what should be measured and how it should be measured – rather than the aims of the analysis. […] techniques such as CBA and CEA are […] defined by measurement rather than economic theory. […] much of the economic evaluation literature gives the label cost-minimisation analysis to what was traditionally called CEA, and specifically restricts the term CEA to choices between alternatives that have similar types of effects but differing levels of effect and costs. […] It can be difficult to specify what the appropriate measure of effect is in CEA. […] care is […] required to ensure that whichever measure of effect is chosen does not mislead or bias the analysis – for example, if one intervention is better at preventing non-fatal heart attacks but is worse at preventing fatal attacks, the choice of effect measure will be crucial.”

“[Health] indicators are usually measures of the value of health, although not usually expressed in money terms. As a result, a third important type of economic evaluation has arisen, called cost–utility analysis (CUA). […] the health measure usually used in CUA is gains in quality-adjusted life years […] it is essentially a composite measure of gains in life expectancy and health-related quality of life. […] the most commonly used practice in CUA is to use the QALY and moreover to assume that each QALY is worth the same irrespective of who gains it and by what route. […] Similarly, CBA in practice focuses on sums of benefits compared to sums of costs, not on the distribution of these between people with different characteristics. It also does not usually take account of whether society places different weights on benefits experienced by different people; for example, there is evidence that many people would prefer health services to put a higher priority on improving the health of younger rather than older people (Tsuchiya et al., 2003).”

“Because CEA does not give a direct comparison between the value of effects and costs, decision rules are far more complex than for CBA and are bounded by restrictions on their applicability. The problem arises when the alternatives being appraised do not have equal costs or benefits, but instead there is a trade-off: the greater benefit that one of the alternatives has is achieved at a higher cost [this is not a rare occurrence, to put it mildly…]. The key problem is how that trade-off is to be represented, and how it can then be interpreted; essentially, encapsulating cost-effectiveness in a single index that can unambiguously be interpreted for decision-making purposes.”

“Although cost-effectiveness analysis can be very useful, its essential inability to help in the kind of choices that cost–benefit analysis allows – an absolute recommendation for a particular activity rather than one contingent on a comparison with alternatives – has proved such a strong limitation that means have been sought to overcome it. The key to this has been the cost-effectiveness threshold or ceiling ratio, which is essentially a level of the CER that any intervention must meet if it is to be regarded as cost-effective. It can also be interpreted as the decision maker’s willingness to pay for a unit of effectiveness. […] One of the problems with this kind of approach is that it is no longer consistent with the conventional aim of CEA. Except under special conditions, it is not consistent with output maximisation constrained by a budget. […] It is useful to distinguish between a comparator that is essentially ‘do nothing about the problem […]’ and one that is ‘another way of doing something about that problem’. The CER that arises from the second of these is […] an incremental cost-effectiveness ratio (ICER) […] in most cases the ICER is the correct measure to use. […] A problem [with using ICERs] is that if only the ICER is evaluated, it must be assumed that the alternative used in the comparator is itself cost-effective; if it is not, the ICER may mislead.”

“The basis of economic costing is […] quite distinct from accounting or financial cost approaches. The process of costing involves three steps: (1) identify and describe the changes in resource use, both increases and decreases, that are associated with the options to be evaluated; (2) quantify those changes in resource use in physical units; and (3) value those resources. […] many markets are not fully competitive. For example, the wages paid to doctors may be a reflection of the lobbying power of medical associations or restrictions to licensing, rather than the value of their skills […] The prices of drugs may reflect the effect of government regulations on licensing, pricing and intellectual property. Deviations of price from opportunity cost may arise from factors such as imperfect competition […] or from distortions to markets created by government interventions. Where these are known, prices should be adjusted […] In practice, such adjustments are difficult to make and would rely on good information on the underlying costs of production, which is often not available. Further, where the perspective is that of the health service, there is an argument for not adjusting prices, on the grounds that the prevailing prices, even if inefficient, are those they must pay and are relevant to their budget. […] Where prices are used, it is important to consider whether the option being evaluated will, if implemented, result in price changes. […] Valuing resource use becomes still more difficult in cases where there are no markets. This includes the value of patients’ time in seeking and receiving care or of caregivers’ time in providing informal supportive care. The latter can be an important element of costs and […] may be particularly important in the evaluation of health care options that rely on such inputs.”

“[A]lthough the emphasis in economic evaluation is on marginal changes in costs and benefits, the available data frequently relate to average costs […] There are two issues with using average cost data. First, the addition to or reduction in costs from increased or decreased resource use may be higher, lower or the same as the average cost. Unfortunately, knowing what the relationship is between average and marginal cost requires information on the latter – the absence of which is the reason average costs are used! Secondly, average cost data obscure potentially important issues with respect to the technical efficiency of providers. If average costs are derived in one setting, for example a hospital, this assumes that the hospital is using the optimal combination of inputs. If average costs are derived from multiple settings, they will include a variety of underlying production technologies and a variety of underlying levels of production efficiency. Average costs are therefore less than ideal, because they comprise a ‘black box’ of underlying cost and production decisions. […] Approaches to costing fall into two broad types: macro- or ‘top-down’ costing, and micro- or ‘bottom-up’ costing […] distinguished largely on the basis of the level of disaggregation […] A top-down approach may involve using pre-existing data on total or average costs and apportioning these in some way to the options being evaluated. […] In contrast, a bottom-up approach identifies, quantifies and values resources in a disaggregated way, so that each element of costs is estimated individually and they are summed up at the end. […] The separation of top-down and bottom-up costing approaches is not always clear. For example, often top-down studies are used to calculate unit costs, which are then combined with resource use data in bottom-up studies.”

“Health care programmes can affect both length and quality of life; these in turn interact with both current and future health care use, relating both to the condition of interest and to other conditions. Weinstein and Stason (1977) argue that the cost of ‘saving’ life in one way should include the future costs to the health service of death from other causes. […] In practice, different analysts respond to this issue in different ways: examples may be found of economic evaluations of mammography screening that do […] and do not […] incorporate future health care costs. Methodological differences of this sort reduce the ability to make valid comparisons between results. In practical terms, this issue is a matter of researcher discretion”.

The stuff included in the last paragraph above is closely linked to stuff covered in the biodemography text I’m currently reading, and I expect to cover related topics in some detail in the future here on the blog. Below a few final observations from the book about discounting:

“It is generally accepted that future costs should be discounted in an economic evaluation and, in CBA, it is also relatively non-controversial that benefits, in monetary terms, should also be discounted. In contrast, there is considerable debate surrounding the issue of whether to discount health outcomes such as QALYs, and what the appropriate discount rate is. […] The debate […] concentrates on the issue of whether people have a time preference for receiving health benefits now rather than in the future in the same way that they might have a time preference for gaining monetary benefits now rather than later in life. Arguments both for and against this view are plausible, and the issue is currently unresolved. […] The effect of not discounting health benefits is to improve the cost-effectiveness of all health care programmes that have benefits beyond the current time period, because not discounting increases the magnitude of the health benefits. But as well as affecting the apparent cost-effectiveness of programmes relative to some benchmark or threshold, the choice of whether to discount will also affect the cost-effectiveness of different health care programmes relative to each other […] Discounting health benefits tends to make those health care programmes with benefits realised mostly in the future, such as prevention, less cost-effective relative to those with benefits realised mostly in the present, such as cure.”

March 5, 2017 Posted by | Books, Economics, health care | Leave a comment


i. “The older I grow, the more I distrust the familiar doctrine that age brings wisdom.” (H. L. Mencken)

ii. “Conscience is the inner voice that warns us somebody may be looking.” (-ll-)

iii. “No matter how happily a woman may be married, it always pleases her to discover that there is a nice man who wishes that she were not.” (-ll-)

iv. “But the whole thing, after all, may be put very simply. I believe that it is better to tell the truth than to lie. I believe that it is better to be free than to be a slave. And I believe that it is better to know than be ignorant.” (-ll-)

v. “What people say, what people do, and what they say they do are entirely different things.” (Margaret Mead)

vi. “It is an open question whether any behavior based on fear of eternal punishment can be regarded as ethical or should be regarded as merely cowardly.” (-ll-)

vii. “Both optimists and pessimists contribute to our society. The optimist invents the airplane and the pessimist the parachute.” (Gladys Bronwyn Stern)

viii. “One thing that’s good about procrastination is that you always have something planned for tomorrow”. (-ll-)

ix. “Orthodoxy is a relaxation of the mind accompanied by a stiffening of the heart.” (Edward Abbey)

x. “When a wise man does not understand, he says: “I do not understand.” The fool and the uncultured are ashamed of their ignorance. They remain silent when a question could bring them wisdom.” (Frank Herbert)

xi. “Every day, no matter how you fight it, you learn a little more about yourself, and all most of it does is teach humility.” (John D. MacDonald)

xii. “No matter what side of the argument you are on, you always find people on your side that you wish were on the other.” (Jascha Heifetz)

xiii. “For a tribe to endure, it must find some way to achieve internal unity—and that way usually is external strife.” (Peter Farb)

xiv. “If swindling pays, then it will not stop. […] you cannot have a good society unless virtue pays.” (Abraham Maslow)

xv. “When we lose the right to be different, we lose the privilege to be free.” (Charles Evans Hughes)

xvi. “Each community has a curious and distorted image of itself which is always flattering.” (Carl Eckart)

xvii. “A scientist’s aim in a discussion with his colleagues is not to persuade, but to clarify.” (Leo Szilard)

xviii. “The man who is too old to learn was probably always too old to learn.” (Henry S. Haskins)

xix. “Many of us are impersonations of what we know we ought to be.” (-ll-)

xx. “The man who feels that he must be hopeful and cheerful to get along ignores the careers of some pretty successful misanthropes.” (-ll-)

March 4, 2017 Posted by | Quotes/aphorisms | Leave a comment

Diabetes and the brain (IV)

Here’s one of my previous posts in the series about the book. In this post I’ll cover material dealing with two acute hyperglycemia-related diabetic complications (DKA and HHS – see below…) as well as multiple topics related to diabetes and stroke. I’ll start out with a few quotes from the book about DKA and HHS:

“DKA [diabetic ketoacidosis] is defined by a triad of hyperglycemia, ketosis, and acidemia and occurs in the absolute or near-absolute absence of insulin. […] DKA accounts for the bulk of morbidity and mortality in children with T1DM. National population-based studies estimate DKA mortality at 0.15% in the United States (4), 0.18–0.25% in Canada (4, 5), and 0.31% in the United Kingdom (6). […] Rates reach 25–67% in those who are newly diagnosed (4, 8, 9). The rates are higher in younger children […] The risk of DKA among patients with pre-existing diabetes is 1–10% annual per person […] DKA can present with mild-to-severe symptoms. […] polyuria and polydipsia […] patients may present with signs of dehydration, such as tachycardia and dry mucus membranes. […] Vomiting, abdominal pain, malaise, and weight loss are common presenting symptoms […] Signs related to the ketoacidotic state include hyperventilation with deep breathing (Kussmaul’s respiration) which is a compensatory respiratory response to an underlying metabolic acidosis. Acetonemia may cause a fruity odor to the breath. […] Elevated glucose levels are almost always present; however, euglycemic DKA has been described (19). Anion-gap metabolic acidosis is the hallmark of this condition and is caused by elevated ketone bodies.”

“Clinically significant cerebral edema occurs in approximately 1% of patients with diabetic ketoacidosis […] DKA-related cerebral edema may represent a continuum. Mild forms resulting in subtle edema may result in modest mental status abnormalities whereas the most severe manifestations result in overt cerebral injury. […] Cerebral edema typically presents 4–12 h after the treatment for DKA is started (28, 29), but can occur at any time. […] Increased intracranial pressure with cerebral edema has been recognized as the leading cause of morbidity and mortality in pediatric patients with DKA (59). Mortality from DKA-related cerebral edema in children is high, up to 90% […] and accounts for 60–90% of the mortality seen in DKA […] many patients are left with major neurological deficits (28, 31, 35).”

“The hyperosmolar hyperglycemic state (HHS) is also an acute complication that may occur in patients with diabetes mellitus. It is seen primarily in patients with T2DM and has previously been referred to as “hyperglycemic hyperosmolar non-ketotic coma” or “hyperglycemic hyperosmolar non-ketotic state” (13). HHS is marked by profound dehydration and hyperglycemia and often by some degree of neurological impairment. The term hyperglycemic hyperosmolar state is used because (1) ketosis may be present and (2) there may be varying degrees of altered sensorium besides coma (13). Like DKA, the basic underlying disorder is inadequate circulating insulin, but there is often enough insulin to inhibit free fatty acid mobilization and ketoacidosis. […] Up to 20% of patients diagnosed with HHS do not have a previous history of diabetes mellitus (14). […] Kitabchi et al. estimated the rate of hospital admissions due to HHS to be lower than DKA, accounting for less than 1% of all primary diabetic admissions (13). […] Glucose levels rise in the setting of relative insulin deficiency. The low levels of circulating insulin prevent lipolysis, ketogenesis, and ketoacidosis (62) but are unable to suppress hyperglycemia, glucosuria, and water losses. […] HHS typically presents with one or more precipitating factors, similar to DKA. […] Acute infections […] account for approximately 32–50% of precipitating causes (13). […] The mortality rates for HHS vary between 10 and 20% (14, 93).”

It should perhaps be noted explicitly that the mortality rates for these complications are particularly high in the settings of either very young individuals (DKA) or in elderly individuals (HHS) who might have multiple comorbidities. Relatedly HHS often develops acutely specifically in settings where the precipitating factor is something really unpleasant like pneumonia or a cardiovascular event, so a high-ish mortality rate is perhaps not that surprising. Nor is it surprising that very young brains are particularly vulnerable in the context of DKA (I already discussed some of the research on these matters in some detail in an earlier post about this book).

This post to some extent covered the topic of ‘stroke in general’, however I wanted to include here also some more data specifically on diabetes-related matters about this topic. Here’s a quote to start off with:

“DM [Diabetes Mellitus] has been consistently shown to represent a strong independent risk factor of ischemic stroke. […] The contribution of hyperglycemia to increased stroke risk is not proven. […] the relationship between hyperglycemia and stroke remains subject of debate. In this respect, the association between hyperglycemia and cerebrovascular disease is established less strongly than the association between hyperglycemia and coronary heart disease. […] The course of stroke in patients with DM is characterized by higher mortality, more severe disability, and higher recurrence rate […] It is now well accepted that the risk of stroke in individuals with DM is equal to that of individuals with a history of myocardial infarction or stroke, but no DM (24–26). This was confirmed in a recently published large retrospective study which enrolled all inhabitants of Denmark (more than 3 million people out of whom 71,802 patients with DM) and were followed-up for 5 years. In men without DM the incidence of stroke was 2.5 in those without and 7.8% in those with prior myocardial infarction, whereas in patients with DM it was 9.6 in those without and 27.4% in those with history of myocardial infarction. In women the numbers were 2.5, 9.0, 10.0, and 14.2%, respectively (22).

That study incidentally is very nice for me in particular to know about, given that I am a Danish diabetic. I do not here face any of the usual tiresome questions about ‘external validity’ and issues pertaining to ‘extrapolating out of sample’ – not only is it quite likely I’ve actually looked at some of the data used in that analysis myself, I also know that I am almost certainly one of the people included in the analysis. Of course you need other data as well to assess risk (e.g. age, see the previously linked post), but this is pretty clean as far as it goes. Moving on…

“The number of deaths from stroke attributable to DM is highest in low-and-middle-income countries […] the relative risk conveyed by DM is greater in younger subjects […] It is not well known whether type 1 or type 2 DM affects stroke risk differently. […] In the large cohort of women enrolled in the Nurses’ Health Study (116,316 women followed for up to 26 years) it was shown that the incidence of total stroke was fourfold higher in women with type 1 DM and twofold higher among women with type 2 DM than for non-diabetic women (33). […] The impact of DM duration as a stroke risk factor has not been clearly defined. […] In this context it is important to note that the actual duration of type 2 DM is difficult to determine precisely […and more generally: “the date of onset of a certain chronic disease is a quantity which is not defined as precisely as mortality“, as Yashin et al. put it – I also talked about this topic in my previous post, but it’s important when you’re looking at these sorts of things and is worth reiterating – US]. […] Traditional risk factors for stroke such as arterial hypertension, dyslipidemia, atrial fibrillation, heart failure, and previous myocardial infarction are more common in people with DM […]. However, the impact of DM on stroke is not just due to the higher prevalence of these risk factors, as the risk of mortality and morbidity remains over twofold increased after correcting for these factors (4, 37). […] It is informative to distinguish between factors that are non-specific and specific to DM. DM-specific factors, including chronic hyperglycemia, DM duration, DM type and complications, and insulin resistance, may contribute to an elevated stroke risk either by amplification of the harmful effect of other “classical” non-specific risk factors, such as hypertension, or by acting independently.”

More than a few variables are known to impact stroke risk, but the fact that many of the risk factors are related to each other (‘fat people often also have high blood pressure’) makes it hard to figure out which variables are most important, how they interact with each other, etc., etc. One might in that context perhaps conceptualize the metabolic syndrome (-MS) as a sort of indicator variable indicating whether a relatively common set of such related potential risk factors of interest are present or not – it is worth noting in that context that the authors include in the text the observation that: “it is yet uncertain if the whole concept of the MS entails more than its individual components. The clustering of risk factors complicates the assessment of the contribution of individual components to the risk of vascular events, as well as assessment of synergistic or interacting effects.” MS confers a two-threefold increased stroke risk, depending on the definition and the population analyzed, so there’s definitely some relevant stuff included in that box, but in the context of developing new treatment options and better assess risk it might be helpful to – to put it simplistically – know if variable X is significantly more important than variable Y (and how the variables interact, etc., etc.). But this sort of information is hard to get.

There’s more than one type of stroke, and the way diabetes modifies the risk of various stroke types is not completely clear:

“Most studies have consistently shown that DM is an important risk factor for ischemic stroke, while the incidence of hemorrhagic stroke in subjects with DM does not seem to be increased. Consequently, the ratio of ischemic to hemorrhagic stroke is higher in patients with DM than in those stroke patients without DM [recall the base rates I’ve mentioned before in the coverage of this book: 80% of strokes are ischemic strokes in Western countries, and 15 % hemorrhagic] […] The data regarding an association between DM and the risk of hemorrhagic stroke are quite conflicting. In the most series no increased risk of cerebral hemorrhage was found (10, 101), and in the Copenhagen Stroke Registry, hemorrhagic stroke was even six times less frequent in diabetic patients than in non-diabetic subjects (102). […] However, in another prospective population-based study DM was associated with an increased risk of primary intracerebral hemorrhage (103). […] The significance of DM as a risk factor of hemorrhagic stroke could differ depending on ethnicity of subjects or type of DM. In the large Nurses’ Health Study type 1 DM increased the risk of hemorrhagic stroke by 3.8 times while type 2 DM did not increase such a risk (96). […] It is yet unclear if DM predominantly predisposes to either large or small vessel ischemic stroke. Nevertheless, lacunar stroke (small, less than 15mm in diameter infarction, cyst-like, frequently multiple) is considered to be the typical type of stroke in diabetic subjects (105–107), and DM may be present in up to 28–43% of patients with cerebral lacunar infarction (108–110).”

The Danish results mentioned above might not be as useful to me as they were before if the type is important, because the majority of those diabetics included were type 2 diabetics. I know from personal experience that it is difficult to type-identify diabetics using the Danish registry data available if you want to work with population-level data, and any type of scheme attempting this will be subject to potentially large misidentification problems. Some subgroups can be presumably correctly identified using diagnostic codes, but a very large number of individuals will be left out of the analyses if you only rely on identification strategies where you’re (at least reasonably?) certain about the type. I’ve worked on these identification problems during my graduate work so perhaps a few more things are worth mentioning here. In the context of diabetic subgroup analyses, misidentification is in general a much larger problem in the context of type 1 results than in the context of type 2 results; unless the study design takes the large prevalence difference of the two conditions into account, the type 1 sample will be much smaller than the type 2 sample in pretty much all analytical contexts, so a small number of misidentified type 2 individuals can have large impacts on the results of the type 1 sample. Type 1s misidentified as type 2 individuals is in general to be expected to be a much smaller problem in terms of the validity of the type 2 analysis; misidentification of that type will cause a loss of power in the context of the type 1 subgroup analysis, which is already low to start with (and it’ll also make the type 1 subgroup analysis even more vulnerable to misidentified type 2s), but it won’t much change the results of the type 2 subgroup analysis in any significant way. Relatedly, even if enough type 2 patients are misidentified to cause problems with the interpretation of the type 1 subgroup analysis, this would not on its own be a good reason to doubt the results of the type 2 subgroup analysis. Another thing to note in terms of these things is that given that misidentification will tend to lead to ‘mixing’, i.e. it’ll make the subgroup results look similar, when outcomes are not similar in the type 1 and the type 2 individuals then this might be taken to be an indicator that something potentially interesting might be going on, because most analyses will struggle with some level of misidentification which will tend to reduce the power of tests of group differences.

What about stroke outcomes? A few observations were included on that topic above, but the book has a lot more stuff on that – some observations on this topic:

“DM is an independent risk factor of death from stroke […]. Tuomilehto et al. (35) calculated that 16% of all stroke mortality in men and 33% in women could be directly attributed to DM. Patients with DM have higher hospital and long-term stroke mortality, more pronounced residual neurological deficits, and more severe disability after acute cerebrovascular accidents […]. The 1-year mortality rate, for example, was twofold higher in diabetic patients compared to non-diabetic subjects (50% vs. 25%) […]. Only 20% of people with DM survive over 5 years after the first stroke and half of these patients die within the first year (36, 128). […] The mechanisms underlying the worse outcome of stroke in diabetic subjects are not fully understood. […] Regarding prevention of stroke in patients with DM, it may be less relevant than in non-DM subjects to distinguish between primary and secondary prevention as all patients with DM are considered to be high-risk subjects regardless of the history of cerebrovascular accidents or the presence of clinical and subclinical vascular lesions. […] The influence of the mode of antihyperglycemic treatment on the risk of stroke is uncertain.

Control of blood pressure is very important in the diabetic setting:

“There are no doubts that there is a linear relation between elevated systolic blood pressure and the risk of stroke, both in people with or without DM. […] Although DM and arterial hypertension represent significant independent risk factors for stroke if they co-occur in the same patient the risk increases dramatically. A prospective study of almost 50 thousand subjects in Finland followed up for 19 years revealed that the hazard ratio for stroke incidence was 1.4, 2.0, 2.5, 3.5, and 4.5 and for stroke mortality was 1.5, 2.6, 3.1, 5.6, and 9.3, respectively, in subjects with an isolated modestly elevated blood pressure (systolic 140–159/diastolic 90–94 mmHg), isolated more severe hypertension (systolic >159 mmHg, diastolic >94 mmHg, or use of antihypertensive drugs), with isolated DM only, with both DM and modestly elevated blood pressure, and with both DM and more severe hypertension, relative to subjects without either of the risk factors (168). […] it remains unclear whether some classes of antihypertensive agents provide a stronger protection against stroke in diabetic patients than others. […] effective antihypertensive treatment is highly beneficial for reduction of stroke risk in diabetic patients, but the advantages of any particular class of antihypertensive medications are not substantially proven.”

Treatment of dyslipidemia is also very important, but here it does seem to matter how you treat it:

“It seems that the beneficial effect of statins is dose-dependent. The lower the LDL level that is achieved the stronger the cardiovascular protection. […] Recently, the results of the meta-analysis of 14 randomized trials of statins in 18,686 patients with DM had been published. It was calculated that statins use in diabetic patients can result in a 21% reduction of the risk of any stroke per 1 mmol/l reduction of LDL achieved […] There is no evidence from trials that supports efficacy of fibrates for stroke prevention in diabetic patients. […] No reduction of stroke risk by fibrates was shown also in a meta-analysis of eight trials enrolled 12,249 patients with type 2 DM (204).”


“Significant reductions in stroke risk in diabetic patients receiving antiplatelet therapy were found in large-scale controlled trials (205). It appears that based on the high incidence of stroke and prevalence of stroke risk factors in the diabetic population the benefits of routine aspirin use for primary and secondary stroke prevention outweigh its potential risk of hemorrhagic stroke especially in patients older than 30 years having at least one additional risk factor (206). […] both guidelines issued by the AHA/ADA or the ESC/EASD on the prevention of cardiovascular disease in patients with DM support the use of aspirin in a dose of 50–325 mg daily for the primary prevention of stroke in subjects older than 40 years of age and additional risk factors, such as DM […] The newer antiplatelet agent, clopidogrel, was more efficacious in prevention of ischemic stroke than aspirin with greater risk reduction in the diabetic cohort especially in those treated with insulin compared to non-diabetics in CAPRIE trial (209). However, the combination of aspirin and clopidogrel does not appear to be more efficacious and safe compared to clopidogrel or aspirin alone”.

When you treat all risk factors aggressively, it turns out that the elevated stroke risk can be substantially reduced. Again the data on this stuff is from Denmark:

“Gaede et al. (216) have shown in the Steno 2 study that intensive multifactorial intervention aimed at correction of hyperglycemia, hypertension, dyslipidemia, and microalbuminuria along with aspirin use resulted in a reduction of cardiovascular morbidity including non-fatal stroke […] recently the results of the extended 13.3 years follow-up of this study were presented and the reduction of cardiovascular mortality by 57% and morbidity by 59% along with the reduction of the number of non-fatal stroke (6 vs. 30 events) in intensively treated group was convincingly demonstrated (217). Antihypertensive, hypolipidemic treatment, use of aspirin should thus be recommended as either primary or secondary prevention of stroke for patients with DM.”

March 3, 2017 Posted by | Books, Cardiology, Diabetes, Epidemiology, Medicine, Neurology, Pharmacology, Statistics | Leave a comment

Biodemography of aging (I)

“The goal of this monograph is to show how questions about the connections between and among aging, health, and longevity can be addressed using the wealth of available accumulated knowledge in the field, the large volumes of genetic and non-genetic data collected in longitudinal studies, and advanced biodemographic models and analytic methods. […] This monograph visualizes aging-related changes in physiological variables and survival probabilities, describes methods, and summarizes the results of analyses of longitudinal data on aging, health, and longevity in humans performed by the group of researchers in the Biodemography of Aging Research Unit (BARU) at Duke University during the past decade. […] the focus of this monograph is studying dynamic relationships between aging, health, and longevity characteristics […] our focus on biodemography/biomedical demography meant that we needed to have an interdisciplinary and multidisciplinary biodemographic perspective spanning the fields of actuarial science, biology, economics, epidemiology, genetics, health services research, mathematics, probability, and statistics, among others.”

The quotes above are from the book‘s preface. In case this aspect was not clear from the comments above, this is the kind of book where you’ll randomly encounter sentences like these:

The simplest model describing negative correlations between competing risks is the multivariate lognormal frailty model. We illustrate the properties of such model for the bivariate case.

“The time-to-event sub-model specifies the latent class-specific expressions for the hazard rates conditional on the vector of biomarkers Yt and the vector of observed covariates X …”

…which means that some parts of the book are really hard to blog; it simply takes more effort to deal with this stuff here than it’s worth. As a result of this my coverage of the book will not provide a remotely ‘balanced view’ of the topics covered in it; I’ll skip a lot of the technical stuff because I don’t think it makes much sense to cover specific models and algorithms included in the book in detail here. However I should probably also emphasize while on this topic that although the book is in general not an easy read, it’s hard to read because ‘this stuff is complicated’, not because the authors are not trying. The authors in fact make it clear already in the preface that some chapters are more easy to read than are others and that some chapters are actually deliberately written as ‘guideposts and way-stations‘, as they put it, in order to make it easier for the reader to find the stuff in which he or she is most interested (“the interested reader can focus directly on the chapters/sections of greatest interest without having to read the entire volume“) – they have definitely given readability aspects some thought, and I very much like the book so far; it’s full of great stuff and it’s very well written.

I have had occasion to question a few of the observations they’ve made, for example I was a bit skeptical about a few of the conclusions they drew in chapter 6 (‘Medical Cost Trajectories and Onset of Age-Associated Diseases’), but this was related to what some would certainly consider to be minor details. In the chapter they describe a model of medical cost trajectories where the post-diagnosis follow-up period is 20 months; this is in my view much too short a follow-up period to draw conclusions about medical cost trajectories in the context of type 2 diabetes, one of the diseases included in the model, which I know because I’m intimately familiar with the literature on that topic; you need to look 7-10 years ahead to get a proper sense of how this variable develops over time – and it really is highly relevant to include those later years, because if you do not you may miss out on a large proportion of the total cost given that a substantial proportion of the total cost of diabetes relate to complications which tend to take some years to develop. If your cost analysis is based on a follow-up period as short as that of that model you may also on a related note draw faulty conclusions about which medical procedures and -subsidies are sensible/cost effective in the setting of these patients, because highly adherent patients may be significantly more expensive in a short run analysis like this one (they show up to their medical appointments and take their medications…) but much cheaper in the long run (…because they take their medications they don’t go blind or develop kidney failure). But as I say, it’s a minor point – this was one condition out of 20 included in the analysis they present, and if they’d addressed all the things that pedants like me might take issue with, the book would be twice as long and it would likely no longer be readable. Relatedly, the model they discuss in that chapter is far from unsalvageable; it’s just that one of the components of interest –  ‘the difference between post- and pre-diagnosis cost levels associated with an acquired comorbidity’ – in the case of at least one disease is highly unlikely to be correct (given the authors’ interpretation of the variable), because there’s some stuff of relevance which the model does not include. I found the model quite interesting, despite the shortcomings, and the results were definitely surprising. (No, the above does not in my opinion count as an example of coverage of a ‘specific model […] in detail’. Or maybe it does, but I included no equations. On reflection I probably can’t promise much more than that, sometimes the details are interesting…)

Anyway, below I’ve added some quotes from the first few chapters of the book and a few remarks along the way.

“The genetics of aging, longevity, and mortality has become the subject of intensive analyses […]. However, most estimates of genetic effects on longevity in GWAS have not reached genome-wide statistical significance (after applying the Bonferroni correction for multiple testing) and many findings remain non-replicated. Possible reasons for slow progress in this field include the lack of a biologically-based conceptual framework that would drive development of statistical models and methods for genetic analyses of data [here I was reminded of Burnham & Anderson’s coverage, in particular their criticism of mindless ‘Let the computer find out’-strategies – the authors of that chapter seem to share their skepticism…], the presence of hidden genetic heterogeneity, the collective influence of many genetic factors (each with small effects), the effects of rare alleles, and epigenetic effects, as well as molecular biological mechanisms regulating cellular functions. […] Decades of studies of candidate genes show that they are not linked to aging-related traits in a straightforward fashion (Finch and Tanzi 1997; Martin 2007). Recent genome-wide association studies (GWAS) have supported this finding by showing that the traits in late life are likely controlled by a relatively large number of common genetic variants […]. Further, GWAS often show that the detected associations are of tiny size (Stranger et al. 2011).”

I think this ties in well with what I’ve previously read on these and related topics – see e.g. the second-last paragraph quoted in my coverage of Richard Alexander’s book, or some of the remarks included in Roberts et al. Anyway, moving on:

“It is well known from epidemiology that values of variables describing physiological states at a given age are associated with human morbidity and mortality risks. Much less well known are the facts that not only the values of these variables at a given age, but also characteristics of their dynamic behavior during the life course are also associated with health and survival outcomes. This chapter [chapter 8 in the book, US] shows that, for monotonically changing variables, the value at age 40 (intercept), the rate of change (slope), and the variability of a physiological variable, at ages 40–60, significantly influence both health-span and longevity after age 60. For non-monotonically changing variables, the age at maximum, the maximum value, the rate of decline after reaching the maximum (right slope), and the variability in the variable over the life course may influence health-span and longevity. This indicates that such characteristics can be important targets for preventive measures aiming to postpone onsets of complex diseases and increase longevity.”

The chapter from which the quotes in the next two paragraphs are taken was completely filled with data from the Framingham Heart Study, and it was hard for me to know what to include here and what to leave out – so you should probably just consider the stuff I’ve included below as samples of the sort of observations included in that part of the coverage.

“To mediate the influence of internal or external factors on lifespan, physiological variables have to show associations with risks of disease and death at different age intervals, or directly with lifespan. For many physiological variables, such associations have been established in epidemiological studies. These include body mass index (BMI), diastolic blood pressure (DBP), systolic blood pressure (SBP), pulse pressure (PP), blood glucose (BG), serum cholesterol (SCH), hematocrit (H), and ventricular rate (VR). […] the connection between BMI and mortality risk is generally J-shaped […] Although all age patterns of physiological indices are non-monotonic functions of age, blood glucose (BG) and pulse pressure (PP) can be well approximated by monotonically increasing functions for both genders. […] the average values of body mass index (BMI) increase with age (up to age 55 for males and 65 for females), and then decline for both sexes. These values do not change much between ages 50 and 70 for males and between ages 60 and 70 for females. […] Except for blood glucose, all average age trajectories of physiological indices differ between males and females. Statistical analysis confirms the significance of these differences. In particular, after age 35 the female BMI increases faster than that of males. […] [When comparing women with less than or equal to 11 years of education [‘LE’] to women with 12 or more years of education [HE]:] The average values of BG for both groups are about the same until age 45. Then the BG curve for the LE females becomes higher than that of the HE females until age 85 where the curves intersect. […] The average values of BMI in the LE group are substantially higher than those among the HE group over the entire age interval. […] The average values of BG for the HE and LE males are very similar […] However, the differences between groups are much smaller than for females.”

They also in the chapter compared individuals with short life-spans [‘SL’, died before the age of 75] and those with long life-spans [‘LL’, 100 longest-living individuals in the relevant sample] to see if the variables/trajectories looked different. They did, for example: “trajectories for the LL females are substantially different from those for the SL females in all eight indices. Specifically, the average values of BG are higher and increase faster in the SL females. The entire age trajectory of BMI for the LL females is shifted to the right […] The average values of DBP [diastolic blood pressure, US] among the SL females are higher […] A particularly notable observation is the shift of the entire age trajectory of BMI for the LL males and females to the right (towards an older age), as compared with the SL group, and achieving its maximum at a later age. Such a pattern is markedly different from that for healthy and unhealthy individuals. The latter is mostly characterized by the higher values of BMI for the unhealthy people, while it has similar ages at maximum for both the healthy and unhealthy groups. […] Physiological aging changes usually develop in the presence of other factors affecting physiological dynamics and morbidity/mortality risks. Among these other factors are year of birth, gender, education, income, occupation, smoking, and alcohol use. An important limitation of most longitudinal studies is the lack of information regarding external disturbances affecting individuals in their day-today life.”

I incidentally noted while I was reading that chapter that a relevant variable ‘lurking in the shadows’ in the context of the male and female BMI trajectories might be changing smoking habits over time; I have not looked at US data on this topic, but I do know that the smoking patterns of Danish males and females during the latter half of the last century were markedly different and changed really quite dramatically in just a few decades; a lot more males than females smoked in the 60es, whereas the proportions of male- and female smokers today are much more similar, because a lot of males have given up smoking (I refer Danish readers to this blog post which I wrote some years ago on these topics). The authors of the chapter incidentally do look a little at data on smokers and they observe that smokers’ BMI are lower than non-smokers (not surprising), and that the smokers’ BMI curve (displaying the relationship between BMI and age) grows at a slower rate than the BMI curve of non-smokers (that this was to be expected is perhaps less clear, at least to me – the authors don’t interpret these specific numbers, they just report them).

The next chapter is one of the chapters in the book dealing with the SEER data I also mentioned not long ago in the context of my coverage of Bueno et al. Some sample quotes from that chapter below:

“To better address the challenge of “healthy aging” and to reduce economic burdens of aging-related diseases, key factors driving the onset and progression of diseases in older adults must be identified and evaluated. An identification of disease-specific age patterns with sufficient precision requires large databases that include various age-specific population groups. Collections of such datasets are costly and require long periods of time. That is why few studies have investigated disease-specific age patterns among older U.S. adults and there is limited knowledge of factors impacting these patterns. […] Information collected in U.S. Medicare Files of Service Use (MFSU) for the entire Medicare-eligible population of older U.S. adults can serve as an example of observational administrative data that can be used for analysis of disease-specific age patterns. […] In this chapter, we focus on a series of epidemiologic and biodemographic characteristics that can be studied using MFSU.”

“Two datasets capable of generating national level estimates for older U.S. adults are the Surveillance, Epidemiology, and End Results (SEER) Registry data linked to MFSU (SEER-M) and the National Long Term Care Survey (NLTCS), also linked to MFSU (NLTCS-M). […] The SEER-M data are the primary dataset analyzed in this chapter. The expanded SEER registry covers approximately 26 % of the U.S. population. In total, the Medicare records for 2,154,598 individuals are available in SEER-M […] For the majority of persons, we have continuous records of Medicare services use from 1991 (or from the time the person reached age 65 after 1990) to his/her death. […] The NLTCS-M data contain two of the six waves of the NLTCS: namely, the cohorts of years 1994 and 1999. […] In total, 34,077 individuals were followed-up between 1994 and 1999. These individuals were given the detailed NLTCS interview […] which has information on risk factors. More than 200 variables were selected”

In short, these data sets are very large, and contain a lot of information. Here are some results/data:

“Among studied diseases, incidence rates of Alzheimer’s disease, stroke, and heart failure increased with age, while the rates of lung and breast cancers, angina pectoris, diabetes, asthma, emphysema, arthritis, and goiter became lower at advanced ages. [..] Several types of age-patterns of disease incidence could be described. The first was a monotonic increase until age 85–95, with a subsequent slowing down, leveling off, and decline at age 100. This pattern was observed for myocardial infarction, stroke, heart failure, ulcer, and Alzheimer’s disease. The second type had an earlier-age maximum and a more symmetric shape (i.e., an inverted U-shape) which was observed for lung and colon cancers, Parkinson’s disease, and renal failure. The majority of diseases (e.g., prostate cancer, asthma, and diabetes mellitus among them) demonstrated a third shape: a monotonic decline with age or a decline after a short period of increased rates. […] The occurrence of age-patterns with a maximum and, especially, with a monotonic decline contradicts the hypothesis that the risk of geriatric diseases correlates with an accumulation of adverse health events […]. Two processes could be operative in the generation of such shapes. First, they could be attributed to the effect of selection […] when frail individuals do not survive to advanced ages. This approach is popular in cancer modeling […] The second explanation could be related to the possibility of under-diagnosis of certain chronic diseases at advanced ages (due to both less pronounced disease symptoms and infrequent doctor’s office visits); however, that possibility cannot be assessed with the available data […this is because the data sets are based on Medicare claims – US]”

“The most detailed U.S. data on cancer incidence come from the SEER Registry […] about 60 % of malignancies are diagnosed in persons aged 65+ years old […] In the U.S., the estimated percent of cancer patients alive after being diagnosed with cancer (in 2008, by current age) was 13 % for those aged 65–69, 25 % for ages 70–79, and 22 % for ages 80+ years old (compared with 40 % of those aged younger than 65 years old) […] Diabetes affects about 21 % of the U.S. population aged 65+ years old (McDonald et al. 2009). However, while more is known about the prevalence of diabetes, the incidence of this disease among older adults is less studied. […] [In multiple previous studies] the incidence rates of diabetes decreased with age for both males and females. In the present study, we find similar patterns […] The prevalence of asthma among the U.S. population aged 65+ years old in the mid-2000s was as high as 7 % […] older patients are more likely to be underdiagnosed, untreated, and hospitalized due to asthma than individuals younger than age 65 […] asthma incidence rates have been shown to decrease with age […] This trend of declining asthma incidence with age is in agreement with our results.”

“The prevalence and incidence of Alzheimer’s disease increase exponentially with age, with the most notable rise occurring through the seventh and eight decades of life (Reitz et al. 2011). […] whereas dementia incidence continues to increase beyond age 85, the rate of increase slows down [which] suggests that dementia diagnosed at advanced ages might be related not to the aging process per se, but associated with age-related risk factors […] Approximately 1–2 % of the population aged 65+ and up to 3–5 % aged 85+ years old suffer from Parkinson’s disease […] There are few studies of Parkinsons disease incidence, especially in the oldest old, and its age patterns at advanced ages remain controversial”.

“One disadvantage of large administrative databases is that certain factors can produce systematic over/underestimation of the number of diagnosed diseases or of identification of the age at disease onset. One reason for such uncertainties is an incorrect date of disease onset. Other sources are latent disenrollment and the effects of study design. […] the date of onset of a certain chronic disease is a quantity which is not defined as precisely as mortality. This uncertainty makes difficult the construction of a unified definition of the date of onset appropriate for population studies.”

“[W]e investigated the phenomenon of multimorbidity in the U.S. elderly population by analyzing mutual dependence in disease risks, i.e., we calculated disease risks for individuals with specific pre-existing conditions […]. In total, 420 pairs of diseases were analyzed. […] For each pair, we calculated age patterns of unconditional incidence rates of the diseases, conditional rates of the second (later manifested) disease for individuals after onset of the first (earlier manifested) disease, and the hazard ratio of development of the subsequent disease in the presence (or not) of the first disease. […] three groups of interrelations were identified: (i) diseases whose risk became much higher when patients had a certain pre-existing (earlier diagnosed) disease; (ii) diseases whose risk became lower than in the general population when patients had certain pre-existing conditions […] and (iii) diseases for which “two-tail” effects were observed: i.e., when the effects are significant for both orders of disease precedence; both effects can be direct (either one of the diseases from a disease pair increases the risk of the other disease), inverse (either one of the diseases from a disease pair decreases the risk of the other disease), or controversial (one disease increases the risk of the other, but the other disease decreases the risk of the first disease from the disease pair). In general, the majority of disease pairs with increased risk of the later diagnosed disease in both orders of precedence were those in which both the pre-existing and later occurring diseases were cancers, and also when both diseases were of the same organ. […] Generally, the effect of dependence between risks of two diseases diminishes with advancing age. […] Identifying mutual relationships in age-associated disease risks is extremely important since they indicate that development of […] diseases may involve common biological mechanisms.”

“in population cohorts, trends in prevalence result from combinations of trends in incidence, population at risk, recovery, and patients’ survival rates. Trends in the rates for one disease also may depend on trends in concurrent diseases, e.g., increasing survival from CHD contributes to an increase in the cancer incidence rate if the individuals who survived were initially susceptible to both diseases.”

March 1, 2017 Posted by | Biology, Books, Cancer/oncology, Cardiology, Demographics, Diabetes, Epidemiology, Genetics, Medicine, Nephrology, Neurology | Leave a comment


i. “Only the most uncritical minds are free from doubt.” (Aldo Leopold)

ii. “If you do not tell the truth about yourself you cannot tell it about other people.” (Virginia Woolf)

iii. “Though we see the same world, we see it through different eyes.” (-ll-)

iv. “No greater mistake can be made than to think that our institutions are fixed or may not be changed for the worse.” (Charles Evans Hughes)

v. “The image of ourselves in the minds of others is the picture of a stranger we shall never see.” (Elizabeth Bibesco)

vi. “Everybody continually tries to get away with as much as he can; and society is a marvelous machine which allows decent people to be cruel without realizing it.” (Émile Chartier)

vii. “When a man steals your wife, there is no better revenge than to let him keep her.” (Sacha Guitry)

viii. “Equipped with his five senses, man explores the universe around him and calls the adventure Science.” (Edwin Hubble)

ix. “There are two kinds of fools: one says, “This is old, therefore it is good”; the other says, “This is new, therefore it is better.” (William Ralph Inge)

x. “We know too many things that are not true.” (Charles Kettering)

xi. “There are truths which one can only say after having won the right to say them.” (Jean Cocteau)

xii. “Where all think alike, no one thinks very much.” (Walter Lippmann)

xiii. “It requires wisdom to understand wisdom: the music is nothing if the audience is deaf.” (-ll-)

xiv. “The past is a foreign country; they do things differently there.” (L.P. Hartley)

xv. “To know is not too demanding: it merely requires memory and time. But to understand is quite a different matter: it requires intellectual ability and training, a self conscious awareness of what one is doing, experience in techniques of analysis and synthesis, and above all, perspective.” (Carroll Quigley)

xvi. “The basis of social relationships is reciprocity: if you cooperate with others, others will cooperate with you.” (-ll-. But be careful…)

xvii. “Self-pity? I see no moral objections to it, the smell drives people away, but that’s a practical objection, and occasionally an advantage.” (E. M. Forster)

xviii. “You are neither right nor wrong because people agree with you.” (Benjamin Graham)

xix. “Men substitute words for reality and then argue about the words.” (Edwin Howard Armstrong)

xx. “Science aims at constructing a world which shall be symbolic of the world of commonplace experience. It is not at all necessary that every individual symbol that is used should represent something in common experience or even something explicable in terms of common experience. The man in the street is always making this demand for concrete explanation of the things referred to in science; but of necessity he must be disappointed. It is like our experience in learning to read. That which is written in a book is symbolic of a story in real life. The whole intention of the book is that ultimately a reader will identify some symbol, say BREAD, with one of the conceptions of familiar life. But it is mischievous to attempt such identifications prematurely, before the letters are strung into words and the words into sentences. The symbol A is not the counterpart of anything in familiar life.” (Arthur Eddington)

February 24, 2017 Posted by | Quotes/aphorisms | Leave a comment

The Ageing Immune System and Health (II)

Here’s the first post about the book. I finished it a while ago but I recently realized I had not completed my intended coverage of the book here on the blog back then, and as some of the book’s material sort-of-kind-of relates to material encountered in a book I’m currently reading (Biodemography of Aging) I decided I might as well finish my coverage of the book now in order to review some things I might have forgot in the meantime, by providing coverage here of some of the material covered in the second half of the book. It’s a nice book with some interesting observations, but as I also pointed out in my first post it is definitely not an easy read. Below I have included some observations from the book’s second half.


“The aged lung is characterised by airspace enlargement similar to, but not identical with acquired emphysema [4]. Such tissue damage is detected even in non-smokers above 50 years of age as the septa of the lung alveoli are destroyed and the enlarged alveolar structures result in a decreased surface for gas exchange […] Additional problems are that surfactant production decreases with age [6] increasing the effort needed to expand the lungs during inhalation in the already reduced thoracic cavity volume where the weakened muscles are unable to thoroughly ventilate. […] As ageing is associated with respiratory muscle strength reduction, coughing becomes difficult making it progressively challenging to eliminate inhaled particles, pollens, microbes, etc. Additionally, ciliary beat frequency (CBF) slows down with age impairing the lungs’ first line of defence: mucociliary clearance [9] as the cilia can no longer repel invading microorganisms and particles. Consequently e.g. bacteria can more easily colonise the airways leading to infections that are frequent in the pulmonary tract of the older adult.”

“With age there are dramatic changes in neutrophil function, including reduced chemotaxis, phagocytosis and bactericidal mechanisms […] reduced bactericidal function will predispose to infection but the reduced chemotaxis also has consequences for lung tissue as this results in increased tissue bystander damage from neutrophil elastases released during migration […] It is currently accepted that alterations in pulmonary PPAR profile, more precisely loss of PPARγ activity, can lead to inflammation, allergy, asthma, COPD, emphysema, fibrosis, and cancer […]. Since it has been reported that PPARγ activity decreases with age, this provides a possible explanation for the increasing incidence of these lung diseases and conditions in older individuals [6].”


“Age is an important risk factor for cancer and subjects aged over 60 also have a higher risk of comorbidities. Approximately 50 % of neoplasms occur in patients older than 70 years […] a major concern for poor prognosis is with cancer patients over 70–75 years. These patients have a lower functional reserve, a higher risk of toxicity after chemotherapy, and an increased risk of infection and renal complications that lead to a poor quality of life. […] [Whereas] there is a difference in organs with higher cancer incidence in developed versus developing countries [,] incidence increases with ageing almost irrespective of country […] The findings from Surveillance, Epidemiology and End Results Program [SEERincidentally I likely shall at some point discuss this one in much more detail, as the aforementioned biodemography textbook covers this data in a lot of detail.. – US] [6] show that almost a third of all cancer are diagnosed after the age of 75 years and 70 % of cancer-related deaths occur after the age of 65 years. […] The traditional clinical trial focus is on younger and healthier patient, i.e. with few or no co-morbidities. These restrictions have resulted in a lack of data about the optimal treatment for older patients [7] and a poor evidence base for therapeutic decisions. […] In the older patient, neutropenia, anemia, mucositis, cardiomyopathy and neuropathy — the toxic effects of chemotherapy — are more pronounced […] The correction of comorbidities and malnutrition can lead to greater safety in the prescription of chemotherapy […] Immunosenescence is a general classification for changes occurring in the immune system during the ageing process, as the distribution and function of cells involved in innate and adaptive immunity are impaired or remodelled […] Immunosenescence is considered a major contributor to cancer development in aged individuals“.

Neurodegenerative diseases:

“Dementia and age-related vision loss are major causes of disability in our ageing population and it is estimated that a third of people aged over 75 are affected. […] age is the largest risk factor for the development of neurodegenerative diseases […] older patients with comorbidities such as atherosclerosis, type II diabetes or those suffering from repeated or chronic systemic bacterial and viral infections show earlier onset and progression of clinical symptoms […] analysis of post-mortem brain tissue from healthy older individuals has provided evidence that the presence of misfolded proteins alone does not correlate with cognitive decline and dementia, implying that additional factors are critical for neural dysfunction. We now know that innate immune genes and life-style contribute to the onset and progression of age-related neuronal dysfunction, suggesting that chronic activation of the immune system plays a key role in the underlying mechanisms that lead to irreversible tissue damage in the CNS. […] Collectively these studies provide evidence for a critical role of inflammation in the pathogenesis of a range of neurodegenerative diseases, but the factors that drive or initiate inflammation remain largely elusive.”

“The effect of infection, mimicked experimentally by administration of bacterial lipopolysaccharide (LPS) has revealed that immune to brain communication is a critical component of a host organism’s response to infection and a collection of behavioural and metabolic adaptations are initiated over the course of the infection with the purpose of restricting the spread of a pathogen, optimising conditions for a successful immune response and preventing the spread of infection to other organisms [10]. These behaviours are mediated by an innate immune response and have been termed ‘sickness behaviours’ and include depression, reduced appetite, anhedonia, social withdrawal, reduced locomotor activity, hyperalgesia, reduced motivation, cognitive impairment and reduced memory encoding and recall […]. Metabolic adaptation to infection include fever, altered dietary intake and reduction in the bioavailability of nutrients that may facilitate the growth of a pathogen such as iron and zinc [10]. These behavioural and metabolic adaptions are evolutionary highly conserved and also occur in humans”.

“Sickness behaviour and transient microglial activation are beneficial for individuals with a normal, healthy CNS, but in the ageing or diseased brain the response to peripheral infection can be detrimental and increases the rate of cognitive decline. Aged rodents exhibit exaggerated sickness and prolonged neuroinflammation in response to systemic infection […] Older people who contract a bacterial or viral infection or experience trauma postoperatively, also show exaggerated neuroinflammatory responses and are prone to develop delirium, a condition which results in a severe short term cognitive decline and a long term decline in brain function […] Collectively these studies demonstrate that peripheral inflammation can increase the accumulation of two neuropathological hallmarks of AD, further strengthening the hypothesis that inflammation i[s] involved in the underlying pathology. […] Studies from our own laboratory have shown that AD patients with mild cognitive impairment show a fivefold increased rate of cognitive decline when contracting a systemic urinary tract or respiratory tract infection […] Apart from bacterial infection, chronic viral infections have also been linked to increased incidence of neurodegeneration, including cytomegalovirus (CMV). This virus is ubiquitously distributed in the human population, and along with other age-related diseases such as cardiovascular disease and cancer, has been associated with increased risk of developing vascular dementia and AD [66, 67].”


“Frailty is associated with changes to the immune system, importantly the presence of a pro-inflammatory environment and changes to both the innate and adaptive immune system. Some of these changes have been demonstrated to be present before the clinical features of frailty are apparent suggesting the presence of potentially modifiable mechanistic pathways. To date, exercise programme interventions have shown promise in the reversal of frailty and related physical characteristics, but there is no current evidence for successful pharmacological intervention in frailty. […] In practice, acute illness in a frail person results in a disproportionate change in a frail person’s functional ability when faced with a relatively minor physiological stressor, associated with a prolonged recovery time […] Specialist hospital services such as surgery [15], hip fractures [16] and oncology [17] have now begun to recognise frailty as an important predictor of mortality and morbidity.

I should probably mention here that this is another area where there’s an overlap between this book and the biodemography text I’m currently reading; chapter 7 of the latter text is about ‘Indices of Cumulative Deficits’ and covers this kind of stuff in a lot more detail than does this one, including e.g. detailed coverage of relevant statistical properties of one such index. Anyway, back to the coverage:

“Population based studies have demonstrated that the incidence of infection and subsequent mortality is higher in populations of frail people. […] The prevalence of pneumonia in a nursing home population is 30 times higher than the general population [39, 40]. […] The limited data available demonstrates that frailty is associated with a state of chronic inflammation. There is also evidence that inflammageing predates a diagnosis of frailty suggesting a causative role. […] A small number of studies have demonstrated a dysregulation of the innate immune system in frailty. Frail adults have raised white cell and neutrophil count. […] High white cell count can predict frailty at a ten year follow up [70]. […] A recent meta-analysis and four individual systematic reviews have found beneficial evidence of exercise programmes on selected physical and functional ability […] exercise interventions may have no positive effect in operationally defined frail individuals. […] To date there is no clear evidence that pharmacological interventions improve or ameliorate frailty.”


“[A]s we get older the time and intensity at which we exercise is severely reduced. Physical inactivity now accounts for a considerable proportion of age-related disease and mortality. […] Regular exercise has been shown to improve neutrophil microbicidal functions which reduce the risk of infectious disease. Exercise participation is also associated with increased immune cell telomere length, and may be related to improved vaccine responses. The anti-inflammatory effect of regular exercise and negative energy balance is evident by reduced inflammatory immune cell signatures and lower inflammatory cytokine concentrations. […] Reduced physical activity is associated with a positive energy balance leading to increased adiposity and subsequently systemic inflammation [5]. […] Elevated neutrophil counts accompany increased inflammation with age and the increased ratio of neutrophils to lymphocytes is associated with many age-related diseases including cancer [7]. Compared to more active individuals, less active and overweight individuals have higher circulating neutrophil counts [8]. […] little is known about the intensity, duration and type of exercise which can provide benefits to neutrophil function. […] it remains unclear whether exercise and physical activity can override the effects of NK cell dysfunction in the old. […] A considerable number of studies have assessed the effects of acute and chronic exercise on measures of T-cell immunesenescence including T cell subsets, phenotype, proliferation, cytokine production, chemotaxis, and co-stimulatory capacity. […] Taken together exercise appears to promote an anti-inflammatory response which is mediated by altered adipocyte function and improved energy metabolism leading to suppression of pro-inflammatory cytokine production in immune cells.”

February 24, 2017 Posted by | Biology, Books, Cancer/oncology, Epidemiology, Immunology, Medicine, Neurology | Leave a comment

Economic Analysis in Healthcare (I)

“This book is written to provide […] a useful balance of theoretical treatment, description of empirical analyses and breadth of content for use in undergraduate modules in health economics for economics students, and for students taking a health economics module as part of their postgraduate training. Although we are writing from a UK perspective, we have attempted to make the book as relevant internationally as possible by drawing on examples, case studies and boxed highlights, not just from the UK, but from a wide range of countries”

I’m currently reading this book. The coverage has been somewhat disappointing because it’s mostly an undergraduate text which has so far mainly been covering concepts and ideas I’m already familiar with, but it’s not terrible – just okay-ish. I have added some observations from the first half of the book below.

“Health economics is the application of economic theory, models and empirical techniques to the analysis of decision making by people, health care providers and governments with respect to health and health care. […] Health economics has evolved into a highly specialised field, drawing on related disciplines including epidemiology, statistics, psychology, sociology, operations research and mathematics […] health economics is not shorthand for health care economics. […] Health economics studies not only the provision of health care, but also how this impacts on patients’ health. Other means by which health can be improved are also of interest, as are the determinants of ill-health. Health economics studies not only how health care affects population health, but also the effects of education, housing, unemployment and lifestyles.”

“Economic analyses have been used to explain the rise in obesity. […] The studies show that reasons for the rise in obesity include: *Technological innovation in food production and transportation that has reduced the cost of food preparation […] *Agricultural innovation and falling food prices that has led to an expansion in food supply […] *A decline in physical activity, both at home and at work […] *An increase in the number of fast-food outlets, resulting in changes to the relative prices of meals […]. *A reduction in the prevalence of smoking, which leads to increases in weight (Chou et al., 2004).”

“[T]he evidence is that ageing is in reality a relatively small factor in rising health care costs. The popular view is known as the ‘expansion of morbidity’ hypothesis. Gruenberg (1977) suggested that the decline in mortality that has led to an increase in the number of older people is because fewer people die from illnesses that they have, rather than because disease incidence and prevalence are lower. Lower mortality is therefore accompanied by greater morbidity and disability. However, Fries (1980) suggested an alternative hypothesis, ‘compression of morbidity’. Lower mortality rates are due to better health amongst the population, so people not only live longer, they are in better health when old. […] Zweifel et al. (1999) examined the hypothesis that the main determinant of high health care costs amongst older people is not the time since they were born, but the time until they die. Their results, confirmed by many subsequent studies, is that proximity to death does indeed explain higher health care costs better than age per se. Seshamani and Gray (2004) estimated that in the UK this is a factor up to 15 years before death, and annual costs increase tenfold during the last 5 years of life. The consensus is that ageing per se contributes little to the continuing rise in health expenditures that all countries face. Much more important drivers are improved quality of care, access to care, and more expensive new technology.”

“The difference between AC [average cost] and MC [marginal cost] is very important in applied health economics. Very often data are available on the average cost of health care services but not on their marginal cost. However, using average costs as if they were marginal costs may mislead. For example, hospital costs will be reduced by schemes that allow some patients to be treated in the community rather than being admitted. Given data on total costs of inpatient stays, it is possible to calculate an average cost per patient. It is tempting to conclude that avoiding an admission will reduce costs by that amount. However, the average includes patients with different levels of illness severity, and the more severe the illness the more costly they will be to treat. Less severely ill patients are most likely to be suitable for treatment in the community, so MC will be lower than AC. Such schemes will therefore produce a lower cost reduction than the estimate of AC suggests.
A problem with multi-product cost functions is that it is not possible to define meaningfully what the AC of a particular product is. If different products share some inputs, the costs of those inputs cannot be solely attributed to any one of them. […] In practice, when multi-product organisations such as hospitals calculate costs for particular products, they use accounting rules to share out the costs of all inputs and calculate average not marginal costs.”

“Studies of economies of scale in the health sector do not give a consistent and generalisable picture. […] studies of scope economies [also] do not show any consistent and generalisable picture. […] The impact of hospital ownership type on a range of key outcomes is generally ambiguous, with different studies yielding conflicting results. […] The association between hospital ownership and patient outcomes is unclear. The evidence is mixed and inconclusive regarding the impact of hospital ownership on access to care, morbidity, mortality, and adverse events.

“Public goods are goods that are consumed jointly by all consumers. The strict economics definition of a public good is that they have two characteristics. The first is non-rivalry. This means that the consumption of a good or service by one person does not prevent anyone else from consuming it. Non-rival goods therefore have large marginal external benefits, which make them socially very desirable but privately unprofitable to provide. Examples of nonrival goods are street lighting and pavements. The second is non-excludability. This means that it is not possible to provide a good or service to one person without letting others also consume it. […] This may lead to a free-rider problem, in which people are unwilling to pay for goods and services that are of value to them. […] Note the distinction between public goods, which are goods and services that are non-rival and non-excludable, and publicly provided goods, which are goods or services that are provided by the government for any reason. […] Most health care products and services are not public goods because they are both rival and excludable. […] However, some health care, particularly public health programmes, does have public good properties.”

“[H]ealth care is typically consumed under conditions of uncertainty with respect to the timing of health care expenditure […] and the amount of expenditure on health care that is required […] The usual solution to such problems is insurance. […] Adverse selection exists when exactly the wrong people, from the point of view of the insurance provider, choose to buy insurance: those with high risks. […] Those who are most likely to buy health insurance are those who have a relatively high probability of becoming ill and maybe also incur greater costs than the average when they are ill. […] Adverse selection arises because of the asymmetry of information between insured and insurer. […] Two approaches are adopted to prevent adverse selection. The first is experience rating, where the insurance provider sets a different insurance premium for different risk groups. Those who apply for health insurance might be asked to undergo a medical examination and
to disclose any relevant facts concerning their risk status.
[…] There are two problems with this approach. First, the cost of acquiring the appropriate information may be high. […] Secondly, it might encourage insurance providers to ‘cherry pick’ people, only choosing to provide insurance to the low risk. This may mean that high-risk people are unable to obtain health insurance at all. […] The second approach is to make health insurance compulsory. […] The problem with this is that low-risk people effectively subsidise the health insurance payments of those with higher risks, which may be regarded […] as inequitable.”

“Health insurance changes the economic incentives facing both the consumers and the providers of health care. One manifestation of these changes is the existence of moral hazard. This is a phenomenon common to all forms of insurance. The suggestion is that when people are insured against risks and their consequences, they are less careful about minimising them. […] Moral hazard arises when it is possible to alter the probability of the insured event, […] or the size of the insured loss […] The extent of the problem depends on the price elasticity of demand […] Three main mechanisms can be used to reduce moral hazard. The first is co-insurance. Many insurance policies require that when an event occurs the insured shares the insured loss […] with the insurer. The co-insurance rate is the percentage of the insured loss that is paid by the insured. The co-payment is the amount that they pay. […] The second is deductibles. A deductible is an amount of money the insured pays when a claim is made irrespective of co-insurance. The insurer will not pay the insured loss unless the deductible is paid by the insured. […] The third is no-claims bonuses. These are payments made by insurers to discourage claims. They usually take the form of reduced insurance premiums in the next period. […] No-claims bonuses typically discourage insurance claims where the payout by the insurer is small.

“The method of reimbursement relates to the way in which health care providers are paid for the services they provide. It is useful to distinguish between reimbursement methods, because they can affect the quantity and quality of health care. […] Retrospective reimbursement at full cost means that hospitals receive payment in full for all health care expenditures incurred in some pre-specified period of time. Reimbursement is retrospective in the sense that not only are hospitals paid after they have provided treatment, but also in that the size of the payment is determined after treatment is provided. […] Which model is used depends on whether hospitals are reimbursed for actual costs incurred, or on a fee-for-service (FFS) basis. […] Since hospital income [in these models] depends on the actual costs incurred (actual costs model) or on the volume of services provided (FFS model) there are few incentives to minimise costs. […] Prospective reimbursement implies that payments are agreed in advance and are not directly related to the actual costs incurred. […] incentives to reduce costs are greater, but payers may need to monitor the quality of care provided and access to services. If the hospital receives the same income regardless of quality, there is a financial incentive to provide low-quality care […] The problem from the point of view of the third-party payer is how best to monitor the activities of health care providers, and how to encourage them to act in a mutually beneficial way. This problem might be reduced if health care providers and third-party payers are linked in some way so that they share common goals. […] Integration between third-party payers and health care providers is a key feature of managed care.

One of the prospective imbursement models applied today may be of particular interest to Danes, as the DRG system is a big part of the financial model of the Danish health care system – so I’ve added a few details about this type of system below:

An example of prospectively set costs per case is the diagnostic-related groups (DRG) pricing scheme introduced into the Medicare system in the USA in 1984, and subsequently used in a number of other countries […] Under this scheme, DRG payments are based on average costs per case in each diagnostic group derived from a sample of hospitals. […] Predicted effects of the DRG pricing scheme are cost shifting, patient shifting and DRG creep. Cost shifting and patient shifting are ways of circumventing the cost-minimising effects of DRG pricing by shifting patients or some of the services provided to patients out of the DRG pricing scheme and into other parts of the system not covered by DRG pricing. For example, instead of being provided on an inpatient basis, treatment might be provided on an outpatient basis where it is reimbursed retrospectively. DRG creep arises when hospitals classify cases into DRGs that carry a higher payment, indicating that they are more complicated than they really are. This might arise, for instance, when cases have multiple diagnoses.”

February 20, 2017 Posted by | Books, Economics, health care | Leave a comment