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)
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).
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)
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’s source coding theorem.
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).
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)).
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.
“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.”
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.”
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.”
“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.”
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).
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.“).
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).
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”).
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.”).
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-)
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 complicated – US] […] (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).”
The great majority of the words below are from books I’ve recently read, I’ve almost not spent any time on vocabulary.com since my last post of this kind; the vocabulary.com guys add new words much too slowly, and most of the words they’ve recently added were not in my opinion all that interesting.
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.”
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-)
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.”
“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).
“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 Parkinson’s 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.”