1. Cognitive Dysfunction in Older Adults With Diabetes: What a Clinician Needs to Know. I’ve talked about these topics before here on the blog (see e.g. these posts on related topics), but this is a good summary article. I have added some observations from the paper below:
“Although cognitive dysfunction is associated with both type 1 and type 2 diabetes, there are several distinct differences observed in the domains of cognition affected in patients with these two types. Patients with type 1 diabetes are more likely to have diminished mental flexibility and slowing of mental speed, whereas learning and memory are largely not affected (8). Patients with type 2 diabetes show decline in executive function, memory, learning, attention, and psychomotor efficiency (9,10).”
“So far, it seems that the risk of cognitive dysfunction in type 2 diabetes may be influenced by glycemic control, hypoglycemia, inflammation, depression, and macro- and microvascular pathology (14). The cumulative impact of these conditions on the vascular etiology may further decrease the threshold at which cognition is affected by other neurological conditions in the aging brain. In patients with type 1 diabetes, it seems as though diabetes has a lesser impact on cognitive dysfunction than those patients with type 2 diabetes. […] Thus, the cognitive decline in patients with type 1 diabetes may be mild and may not interfere with their functionality until later years, when other aging-related factors become important. […] However, recent studies have shown a higher prevalence of cognitive dysfunction in older patients (>60 years of age) with type 1 diabetes (5).”
“Unlike other chronic diseases, diabetes self-care involves many behaviors that require various degrees of cognitive pliability and insight to perform proper self-care coordination and planning. Glucose monitoring, medications and/or insulin injections, pattern management, and diet and exercise timing require participation from different domains of cognitive function. In addition, the recognition, treatment, and prevention of hypoglycemia, which are critical for the older population, also depend in large part on having intact cognition.
The reason a clinician needs to recognize different domains of cognition affected in patients with diabetes is to understand which self-care behavior will be affected in that individual. […] For example, a patient with memory problems may forget to take insulin doses, forget to take medications/insulin on time, or forget to eat on time. […] Cognitively impaired patients using insulin are more likely to not know what to do in the event of low blood glucose or how to manage medication on sick days (34). Patients with diminished mental flexibility and processing speed may do well with a simple regimen but may fail if the regimen is too complex. In general, older patients with diabetes with cognitive dysfunction are less likely to be involved in diabetes self-care and glucose monitoring compared with age-matched control subjects (35). […] Other comorbidities associated with aging and diabetes also add to the burden of cognitive impairment and its impact on self-care abilities. For example, depression is associated with a greater decline in cognitive function in patients with type 2 diabetes (36). Depression also can independently negatively impact the motivation to practice self-care.”
“Recently, there is an increasing discomfort with the use of A1C as a sole parameter to define glycemic goals in the older population. Studies have shown that A1C values in the older population may not reflect the same estimated mean glucose as in the younger population. Possible reasons for this discrepancy are the commonly present comorbidities that impact red cell life span (e.g., anemia, uremia, renal dysfunction, blood transfusion, erythropoietin therapy) (45,46). In addition, A1C level does not reflect glucose excursions and variability. […] Thus, it is prudent to avoid A1C as the sole measure of glycemic goal in this population. […] In patients who need insulin therapy, simplification, also known as de-intensification of the regimen, is generally recommended in all frail patients, especially if they have cognitive dysfunction (37,49). However, the practice has not caught up with the recommendations as shown by large observational studies showing unnecessary intensive control in patients with diabetes and dementia (50–52).”
“With advances in the past few decades, we now see a larger number of patients with type 1 diabetes who are aging successfully and facing the new challenges that aging brings. […] Patients with type 1 diabetes are typically proactive in their disease management and highly disciplined. Cognitive dysfunction in these patients creates significant distress for the first time in their lives; they suddenly feel a “lack of control” over the disease they have managed for many decades. The addition of autonomic dysfunction, gastropathy, or neuropathy may result in wider glucose excursions. These patients are usually more afraid of hyperglycemia than hypoglycemia — both of which they have managed for many years. However, cognitive dysfunction in older adults with type 1 diabetes has been found to be associated with hypoglycemic unawareness and glucose variability (5), which in turn increases the risk of severe hypoglycemia (54). The need for goal changes to avoid hypoglycemia and accept some hyperglycemia can be very difficult for many of these patients.”
“From 2006 to 2013, use increased for metformin (from 47.6 to 53.5%), dipeptidyl peptidase 4 inhibitors (0.5 to 14.9%), and insulin (17.1 to 23.0%) but declined for sulfonylureas (38.8 to 30.8%) and thiazolidinediones (28.5 to 5.6%; all P < 0.001). […] The overall rate of severe hypoglycemia remained the same (1.3 per 100 person-years; P = 0.72), declined modestly among the oldest patients (from 2.9 to 2.3; P < 0.001), and remained high among those with two or more comorbidities (3.2 to 3.5; P = 0.36). […] During the recent 8-year period, the use of glucose-lowering drugs has changed dramatically among patients with T2DM. […] The use of older classes of medications, such as sulfonylureas and thiazolidinediones, declined. During this time, glycemic control of T2DM did not improve in the overall population and remained poor among nearly a quarter of the youngest patients. Rates of severe hypoglycemia remained largely unchanged, with the oldest patients and those with multiple comorbidities at highest risk. These findings raise questions about the value of the observed shifts in drug utilization toward newer and costlier medications.”
“Our findings are consistent with a prior study of drug prescribing in U.S. ambulatory practice conducted from 1997 to 2012 (2). In that study, similar increases in DPP-4 inhibitor and insulin analog prescribing were observed; these changes were accompanied by a 61% increase in drug expenditures (2). Our study extends these findings to drug utilization and demonstrates that these increases occurred in all age and comorbidity subgroups. […] In contrast, metformin use increased only modestly between 2006 and 2013 and remained relatively low among older patients and those with two or more comorbidities. Although metformin is recommended as first-line therapy (26), it may be underutilized as the initial agent for the treatment of T2DM (27). Its use may be additionally limited by coexisting contraindications, such as chronic kidney disease (28).”
“The proportion of patients with a diagnosis of diabetes who did not fill any glucose-lowering medications declined slightly (25.7 to 24.1%; P < 0.001).”
That is, one in four people who had a diagnosis of type 2 diabetes were not taking any prescription drugs for their health condition. I wonder how many of those people have read wikipedia articles like this one…
“When considering treatment complexity, the use of oral monotherapy increased slightly (from 24.3 to 26.4%) and the use of multiple (two or more) oral agents declined (from 33.0 to 26.5%), whereas the use of insulin alone and in combination with oral agents increased (from 6.0 to 8.5% and from 11.1 to 14.6%, respectively; all P values <0.001).”
“Between 1987 and 2011, per person medical spending attributable to diabetes doubled (4). More than half of the increase was due to prescription drug spending (4). Despite these spending increases and greater utilization of newly developed medications, we showed no concurrent improvements in overall glycemic control or the rates of severe hypoglycemia in our study. Although the use of newer and more expensive agents may have other important benefits (44), further studies are needed to define the value and cost-effectiveness of current treatment options.”
“Using the 2011–2013 Medical Expenditure Panel Survey, bivariate and regression analyses were conducted to compare demographic characteristics, medical service use, diabetes care, and health status among privately insured adult respondents with diabetes, aged 18–64 years (N = 1,461) by lower (<200% of the federal poverty level) and higher (≥200% of the federal poverty level) income and deductible vs. no deductible (ND), low deductible ($1,000/$2,400) (LD), and high deductible (>$1,000/$2,400) (HD). The National Health Interview Survey 2012–2014 was used to analyze differences in medical debt and delayed/avoided needed care among adult respondents with diabetes (n = 4,058) by income. […] Compared with privately insured respondents with diabetes with ND, privately insured lower-income respondents with diabetes with an LD report significant decreases in service use for primary care, checkups, and specialty visits (27%, 39%, and 77% lower, respectively), and respondents with an HD decrease use by 42%, 65%, and 86%, respectively. Higher-income respondents with an LD report significant decreases in specialty (28%) and emergency department (37%) visits.”
“The reduction in ambulatory visits made by lower-income respondents with ND compared with lower-income respondents with an LD or HD is far greater than for higher-income patients. […] The substantial reduction in checkup (preventive) and specialty visits by those with a lower income who have an HDHP [high-deductible health plan, US] implies a very different pattern of service use compared with lower-income respondents who have ND and with higher-income respondents. Though preventive visits require no out-of-pocket costs, reduced preventive service use with HDHPs is well established and might be the result of patients being unaware of this benefit or their concern about findings that could lead to additional expenses (31). Such sharply reduced service use by low-income respondents with diabetes may not be desirable. Patients with diabetes benefit from assessment of diabetes control, encouragement and reinforcement of behavior change and medication use, and early detection and treatment of diabetes complications or concomitant disease.”
“OBJECTIVE To study long-term mortality, causes of death, and end-stage renal disease (ESRD) in people diagnosed with type 1 diabetes at age 15–29 years.
RESEARCH DESIGN AND METHODS This nationwide, population-based cohort with type 1 diabetes diagnosed during 1978–1982 (n = 719) was followed from diagnosis until death, emigration, or September 2013. Linkages to the Norwegian Cause of Death Registry and the Norwegian Renal Registry provided information on causes of death and whether ESRD was present.
RESULTS During 30 years’ follow-up, 4.6% of participants developed ESRD and 20.6% (n = 148; 106 men and 42 women) died. Cumulative mortality by years since diagnosis was 6.0% (95% CI 4.5–8.0) at 10 years, 12.2% (10.0–14.8) at 20 years, and 18.4% (15.8–21.5) at 30 years. The SMR [standardized mortality ratio] was 4.4 (95% CI 3.7–5.1). Mean time from diagnosis of diabetes to ESRD was 23.6 years (range 14.2–33.5). Death was caused by chronic complications (32.2%), acute complications (20.5%), violent death (19.9%), or any other cause (27.4%). Death was related to alcohol in 15% of cases. SMR for alcohol-related death was 6.8 (95% CI 4.5–10.3), for cardiovascular death was 7.3 (5.4–10.0), and for violent death was 3.6 (2.3–5.3).
CONCLUSIONS The cumulative incidence of ESRD was low in this cohort with type 1 diabetes followed for 30 years. Mortality was 4.4 times that of the general population, and more than 50% of all deaths were caused by acute or chronic complications. A relatively high proportion of deaths were related to alcohol.”
Some additional observations from the paper:
“Studies assessing causes of death in type 1 diabetes are most frequently conducted in individuals diagnosed during childhood (1–7) or without evaluating the effect of age at diagnosis (8,9). Reports on causes of death in cohorts of patients diagnosed during late adolescence or young adulthood, with long-term follow-up, are less frequent (10). […] Adherence to treatment during this age is poor and the risk of acute diabetic complications is high (13–16). Mortality may differ between those with diabetes diagnosed during this period of life and those diagnosed during childhood.”
“Mortality was between four and five times higher than in the general population […]. The excess mortality was similar for men […] and women […]. SMR was higher in the lower age bands — 6.7 (95% CI 3.9–11.5) at 15–24 years and 7.3 (95% CI 5.2–10.1) at 25–34 years — compared with the higher age bands: 3.7 (95% CI 2.7–4.9) at 45–54 years and 3.9 (95% CI 2.6–5.8) at 55–65 years […]. The Cox regression model showed that the risk of death increased significantly by age at diagnosis (HR 1.1; 95% CI 1.1–1.2; P < 0.001) and was eight to nine times higher if ESRD was present (HR 8.7; 95% CI 4.8–15.5; P < 0.0001). […] the underlying cause of death was diabetes in 57 individuals (39.0%), circulatory in 22 (15.1%), cancer in 18 (12.3%), accidents or intoxications in 20 (13.7%), suicide in 8 (5.5%), and any other cause in 21 (14.4%) […] In addition, diabetes contributed to death in 29.5% (n = 43) and CVD contributed to death in 10.9% (n = 29) of the 146 cases. Diabetes was mentioned on the death certificate for 68.2% of the cohort but for only 30.0% of the violent deaths. […] In 60% (88/146) of the cases the review committee considered death to be related to diabetes, whereas in 40% (58/146) the cause was unrelated to diabetes or had an unknown relation to diabetes. According to the clinical committee, acute complications caused death in 20.5% (30/146) of the cases; 20 individuals died as a result of DKA and 10 from hypoglycemia. […] Chronic complications caused the largest proportion of deaths (47/146; 32.2%) and increased with increasing duration of diabetes […]. Among individuals dying as a result of chronic complications (n = 47), CVD caused death in 94% (n = 44) and renal failure in 6% (n = 3). ESRD contributed to death in 22.7% (10/44) of those dying from CVD. Cardiovascular death occurred at mortality rates seven times higher than those in the general population […]. ESRD caused or contributed to death in 13 of 14 cases, when present.”
“Violence (intoxications, accidents, and suicides) was the leading cause of death before 10 years’ duration of diabetes; thereafter it was only a minor cause […] Insulin was used in two of seven suicides. […] According to the available medical records and autopsy reports, about 20% (29/146) of the deceased misused alcohol. In 15% (22/146) alcohol-related ICD-10 codes were listed on the death certificate (18% [19/106] of men and 8% [3/40] of women). In 10 cases the cause of death was uncertain but considered to be related to alcohol or diabetes […] The SMR for alcohol-related death was high when considering the underlying cause of death (5.0; 95% CI 2.5–10.0), and even higher when considering all alcohol-related ICD-10 codes listed on the death certificate (6.8; 95% CI 4.5–10.3). The cause of death was associated with alcohol in 21.8% (19/87) of those who died with less than 20 years’ diabetes duration. Drug abuse was noted on the death certificate in only two cases.”
“During follow-up, 33 individuals (4.6%; 22 men and 11 women) developed ESRD as a result of diabetic nephropathy. Mean time from diagnosis of diabetes to ESRD was 23.6 years (range 14.2–33.5 years). Cumulative incidence of ESRD by years since diagnosis of diabetes was 1.4% (95% CI 0.7–2.7) at 20 years and 4.8% (95% CI 3.4–6.9) at 30 years.”
“This study highlights three important findings. First, among individuals who were diagnosed with type 1 diabetes in late adolescence and early adulthood and had good access to health care, and who were followed for 30 years, mortality was four to five times that of the general population. Second, 15% of all deaths were associated with alcohol, and the SMR for alcohol-related deaths was 6.8. Third, there was a relatively low cumulative incidence of ESRD (4.8%) 30 years after the diagnosis of diabetes.
We report mortality higher than those from a large, population-based study from Finland that found cumulative mortality around 6% at 20 years’ and 15% at 30 years’ duration of diabetes among a population with age at onset and year of diagnosis similar to those in our cohort (10). The corresponding numbers in our cohort were 12% and 18%, respectively; the discrepancy was particularly high at 20 years. The SMR in the Finnish cohort was lower than that in our cohort (2.6–3.0 vs. 3.7–5.1), and those authors reported the SMR to be lower in late-onset diabetes (at age 15–29 years) compared with early-onset diabetes (at age 23). The differences between the Norwegian and Finnish data are difficult to explain since both reports are from countries with good access to health care and a high incidence of type 1 diabetes.”
However the reason for the somewhat different SMRs in these two reasonably similar countries may actually be quite simple – the important variable may be alcohol:
“Finland and Norway are appropriate to compare because they share important population and welfare characteristics. There are, however, significant differences in drinking levels and alcohol-related mortality: the Finnish population consumes more alcohol and the Norwegian population consumes less. The mortality rates for deaths related to alcohol are about three to four times higher in Finland than in Norway (30). […] The markedly higher SMR in our cohort can probably be explained by the lower mortality rates for alcohol-related mortality in the general population. […] In conclusion, the high mortality reported in this cohort with an onset of diabetes in late adolescence and young adulthood draws attention to people diagnosed during a vulnerable period of life. Both acute and chronic complications cause substantial premature mortality […] Our study suggests that increased awareness of alcohol-related death should be encouraged in clinics providing health care to this group of patients.”
My working assumption as I was reading part two of the book was that I would not be covering that part of the book in much detail here because it would simply be too much work to make such posts legible to the readership of this blog. However I then later, while writing this post, had the thought that given that almost nobody reads along here anyway (I’m not complaining, mind you – this is how I like it these days), the main beneficiary of my blog posts will always be myself, which lead to the related observation/notion that I should not be limiting my coverage of interesting stuff here simply because some hypothetical and probably nonexistent readership out there might not be able to follow the coverage. So when I started out writing this post I was working under the assumption that it would be my last post about the book, but I now feel sure that if I find the time I’ll add at least one more post about the book’s statistics coverage. On a related note I am explicitly making the observation here that this post was written for my benefit, not yours. You can read it if you like, or not, but it was not really written for you.
I have added bold a few places to emphasize key concepts and observations from the quoted paragraphs and in order to make the post easier for me to navigate later (all the italics below are on the other hand those of the authors of the book).
“Biodemography is a multidisciplinary branch of science that unites under its umbrella various analytic approaches aimed at integrating biological knowledge and methods and traditional demographic analyses to shed more light on variability in mortality and health across populations and between individuals. Biodemography of aging is a special subfield of biodemography that focuses on understanding the impact of processes related to aging on health and longevity.”
“Mortality rates as a function of age are a cornerstone of many demographic analyses. The longitudinal age trajectories of biomarkers add a new dimension to the traditional demographic analyses: the mortality rate becomes a function of not only age but also of these biomarkers (with additional dependence on a set of sociodemographic variables). Such analyses should incorporate dynamic characteristics of trajectories of biomarkers to evaluate their impact on mortality or other outcomes of interest. Traditional analyses using baseline values of biomarkers (e.g., Cox proportional hazards or logistic regression models) do not take into account these dynamics. One approach to the evaluation of the impact of biomarkers on mortality rates is to use the Cox proportional hazards model with time-dependent covariates; this approach is used extensively in various applications and is available in all popular statistical packages. In such a model, the biomarker is considered a time-dependent covariate of the hazard rate and the corresponding regression parameter is estimated along with standard errors to make statistical inference on the direction and the significance of the effect of the biomarker on the outcome of interest (e.g., mortality). However, the choice of the analytic approach should not be governed exclusively by its simplicity or convenience of application. It is essential to consider whether the method gives meaningful and interpretable results relevant to the research agenda. In the particular case of biodemographic analyses, the Cox proportional hazards model with time-dependent covariates is not the best choice.”
“Longitudinal studies of aging present special methodological challenges due to inherent characteristics of the data that need to be addressed in order to avoid biased inference. The challenges are related to the fact that the populations under study (aging individuals) experience substantial dropout rates related to death or poor health and often have co-morbid conditions related to the disease of interest. The standard assumption made in longitudinal analyses (although usually not explicitly mentioned in publications) is that dropout (e.g., death) is not associated with the outcome of interest. While this can be safely assumed in many general longitudinal studies (where, e.g., the main causes of dropout might be the administrative end of the study or moving out of the study area, which are presumably not related to the studied outcomes), the very nature of the longitudinal outcomes (e.g., measurements of some physiological biomarkers) analyzed in a longitudinal study of aging assumes that they are (at least hypothetically) related to the process of aging. Because the process of aging leads to the development of diseases and, eventually, death, in longitudinal studies of aging an assumption of non-association of the reason for dropout and the outcome of interest is, at best, risky, and usually is wrong. As an illustration, we found that the average trajectories of different physiological indices of individuals dying at earlier ages markedly deviate from those of long-lived individuals, both in the entire Framingham original cohort […] and also among carriers of specific alleles […] In such a situation, panel compositional changes due to attrition affect the averaging procedure and modify the averages in the total sample. Furthermore, biomarkers are subject to measurement error and random biological variability. They are usually collected intermittently at examination times which may be sparse and typically biomarkers are not observed at event times. It is well known in the statistical literature that ignoring measurement errors and biological variation in such variables and using their observed “raw” values as time-dependent covariates in a Cox regression model may lead to biased estimates and incorrect inferences […] Standard methods of survival analysis such as the Cox proportional hazards model (Cox 1972) with time-dependent covariates should be avoided in analyses of biomarkers measured with errors because they can lead to biased estimates.”
“Statistical methods aimed at analyses of time-to-event data jointly with longitudinal measurements have become known in the mainstream biostatistical literature as “joint models for longitudinal and time-to-event data” (“survival” or “failure time” are often used interchangeably with “time-to-event”) or simply “joint models.” This is an active and fruitful area of biostatistics with an explosive growth in recent years. […] The standard joint model consists of two parts, the first representing the dynamics of longitudinal data (which is referred to as the “longitudinal sub-model”) and the second one modeling survival or, generally, time-to-event data (which is referred to as the “survival sub-model”). […] Numerous extensions of this basic model have appeared in the joint modeling literature in recent decades, providing great flexibility in applications to a wide range of practical problems. […] The standard parameterization of the joint model (11.2) assumes that the risk of the event at age t depends on the current “true” value of the longitudinal biomarker at this age. While this is a reasonable assumption in general, it may be argued that additional dynamic characteristics of the longitudinal trajectory can also play a role in the risk of death or onset of a disease. For example, if two individuals at the same age have exactly the same level of some biomarker at this age, but the trajectory for the first individual increases faster with age than that of the second one, then the first individual can have worse survival chances for subsequent years. […] Therefore, extensions of the basic parameterization of joint models allowing for dependence of the risk of an event on such dynamic characteristics of the longitudinal trajectory can provide additional opportunities for comprehensive analyses of relationships between the risks and longitudinal trajectories. Several authors have considered such extended models. […] joint models are computationally intensive and are sometimes prone to convergence problems [however such] models provide more efficient estimates of the effect of a covariate […] on the time-to-event outcome in the case in which there is […] an effect of the covariate on the longitudinal trajectory of a biomarker. This means that analyses of longitudinal and time-to-event data in joint models may require smaller sample sizes to achieve comparable statistical power with analyses based on time-to-event data alone (Chen et al. 2011).”
“To be useful as a tool for biodemographers and gerontologists who seek biological explanations for observed processes, models of longitudinal data should be based on realistic assumptions and reflect relevant knowledge accumulated in the field. An example is the shape of the risk functions. Epidemiological studies show that the conditional hazards of health and survival events considered as functions of risk factors often have U- or J-shapes […], so a model of aging-related changes should incorporate this information. In addition, risk variables, and, what is very important, their effects on the risks of corresponding health and survival events, experience aging-related changes and these can differ among individuals. […] An important class of models for joint analyses of longitudinal and time-to-event data incorporating a stochastic process for description of longitudinal measurements uses an epidemiologically-justified assumption of a quadratic hazard (i.e., U-shaped in general and J-shaped for variables that can take values only on one side of the U-curve) considered as a function of physiological variables. Quadratic hazard models have been developed and intensively applied in studies of human longitudinal data”.
“Various approaches to statistical model building and data analysis that incorporate unobserved heterogeneity are ubiquitous in different scientific disciplines. Unobserved heterogeneity in models of health and survival outcomes can arise because there may be relevant risk factors affecting an outcome of interest that are either unknown or not measured in the data. Frailty models introduce the concept of unobserved heterogeneity in survival analysis for time-to-event data. […] Individual age trajectories of biomarkers can differ due to various observed as well as unobserved (and unknown) factors and such individual differences propagate to differences in risks of related time-to-event outcomes such as the onset of a disease or death. […] The joint analysis of longitudinal and time-to-event data is the realm of a special area of biostatistics named “joint models for longitudinal and time-to-event data” or simply “joint models” […] Approaches that incorporate heterogeneity in populations through random variables with continuous distributions (as in the standard joint models and their extensions […]) assume that the risks of events and longitudinal trajectories follow similar patterns for all individuals in a population (e.g., that biomarkers change linearly with age for all individuals). Although such homogeneity in patterns can be justifiable for some applications, generally this is a rather strict assumption […] A population under study may consist of subpopulations with distinct patterns of longitudinal trajectories of biomarkers that can also have different effects on the time-to-event outcome in each subpopulation. When such subpopulations can be defined on the base of observed covariate(s), one can perform stratified analyses applying different models for each subpopulation. However, observed covariates may not capture the entire heterogeneity in the population in which case it may be useful to conceive of the population as consisting of latent subpopulations defined by unobserved characteristics. Special methodological approaches are necessary to accommodate such hidden heterogeneity. Within the joint modeling framework, a special class of models, joint latent class models, was developed to account for such heterogeneity […] The joint latent class model has three components. First, it is assumed that a population consists of a fixed number of (latent) subpopulations. The latent class indicator represents the latent class membership and the probability of belonging to the latent class is specified by a multinomial logistic regression function of observed covariates. It is assumed that individuals from different latent classes have different patterns of longitudinal trajectories of biomarkers and different risks of event. The key assumption of the model is conditional independence of the biomarker and the time-to-events given the latent classes. Then the class-specific models for the longitudinal and time-to-event outcomes constitute the second and third component of the model thus completing its specification. […] the latent class stochastic process model […] provides a useful tool for dealing with unobserved heterogeneity in joint analyses of longitudinal and time-to-event outcomes and taking into account hidden components of aging in their joint influence on health and longevity. This approach is also helpful for sensitivity analyses in applications of the original stochastic process model. We recommend starting the analyses with the original stochastic process model and estimating the model ignoring possible hidden heterogeneity in the population. Then the latent class stochastic process model can be applied to test hypotheses about the presence of hidden heterogeneity in the data in order to appropriately adjust the conclusions if a latent structure is revealed.”
“The longitudinal genetic-demographic model (or the genetic-demographic model for longitudinal data) […] combines three sources of information in the likelihood function: (1) follow-up data on survival (or, generally, on some time-to-event) for genotyped individuals; (2) (cross-sectional) information on ages at biospecimen collection for genotyped individuals; and (3) follow-up data on survival for non-genotyped individuals. […] Such joint analyses of genotyped and non-genotyped individuals can result in substantial improvements in statistical power and accuracy of estimates compared to analyses of the genotyped subsample alone if the proportion of non-genotyped participants is large. Situations in which genetic information cannot be collected for all participants of longitudinal studies are not uncommon. They can arise for several reasons: (1) the longitudinal study may have started some time before genotyping was added to the study design so that some initially participating individuals dropped out of the study (i.e., died or were lost to follow-up) by the time of genetic data collection; (2) budget constraints prohibit obtaining genetic information for the entire sample; (3) some participants refuse to provide samples for genetic analyses. Nevertheless, even when genotyped individuals constitute a majority of the sample or the entire sample, application of such an approach is still beneficial […] The genetic stochastic process model […] adds a new dimension to genetic biodemographic analyses, combining information on longitudinal measurements of biomarkers available for participants of a longitudinal study with follow-up data and genetic information. Such joint analyses of different sources of information collected in both genotyped and non-genotyped individuals allow for more efficient use of the research potential of longitudinal data which otherwise remains underused when only genotyped individuals or only subsets of available information (e.g., only follow-up data on genotyped individuals) are involved in analyses. Similar to the longitudinal genetic-demographic model […], the benefits of combining data on genotyped and non-genotyped individuals in the genetic SPM come from the presence of common parameters describing characteristics of the model for genotyped and non-genotyped subsamples of the data. This takes into account the knowledge that the non-genotyped subsample is a mixture of carriers and non-carriers of the same alleles or genotypes represented in the genotyped subsample and applies the ideas of heterogeneity analyses […] When the non-genotyped subsample is substantially larger than the genotyped subsample, these joint analyses can lead to a noticeable increase in the power of statistical estimates of genetic parameters compared to estimates based only on information from the genotyped subsample. This approach is applicable not only to genetic data but to any discrete time-independent variable that is observed only for a subsample of individuals in a longitudinal study.”
“Despite an existing tradition of interpreting differences in the shapes or parameters of the mortality rates (survival functions) resulting from the effects of exposure to different conditions or other interventions in terms of characteristics of individual aging, this practice has to be used with care. This is because such characteristics are difficult to interpret in terms of properties of external and internal processes affecting the chances of death. An important question then is: What kind of mortality model has to be developed to obtain parameters that are biologically interpretable? The purpose of this chapter is to describe an approach to mortality modeling that represents mortality rates in terms of parameters of physiological changes and declining health status accompanying the process of aging in humans. […] A traditional (demographic) description of changes in individual health/survival status is performed using a continuous-time random Markov process with a finite number of states, and age-dependent transition intensity functions (transitions rates). Transitions to the absorbing state are associated with death, and the corresponding transition intensity is a mortality rate. Although such a description characterizes connections between health and mortality, it does not allow for studying factors and mechanisms involved in the aging-related health decline. Numerous epidemiological studies provide compelling evidence that health transition rates are influenced by a number of factors. Some of them are fixed at the time of birth […]. Others experience stochastic changes over the life course […] The presence of such randomly changing influential factors violates the Markov assumption, and makes the description of aging-related changes in health status more complicated. […] The age dynamics of influential factors (e.g., physiological variables) in connection with mortality risks has been described using a stochastic process model of human mortality and aging […]. Recent extensions of this model have been used in analyses of longitudinal data on aging, health, and longevity, collected in the Framingham Heart Study […] This model and its extensions are described in terms of a Markov stochastic process satisfying a diffusion-type stochastic differential equation. The stochastic process is stopped at random times associated with individuals’ deaths. […] When an individual’s health status is taken into account, the coefficients of the stochastic differential equations become dependent on values of the jumping process. This dependence violates the Markov assumption and renders the conditional Gaussian property invalid. So the description of this (continuously changing) component of aging-related changes in the body also becomes more complicated. Since studying age trajectories of physiological states in connection with changes in health status and mortality would provide more realistic scenarios for analyses of available longitudinal data, it would be a good idea to find an appropriate mathematical description of the joint evolution of these interdependent processes in aging organisms. For this purpose, we propose a comprehensive model of human aging, health, and mortality in which the Markov assumption is fulfilled by a two-component stochastic process consisting of jumping and continuously changing processes. The jumping component is used to describe relatively fast changes in health status occurring at random times, and the continuous component describes relatively slow stochastic age-related changes of individual physiological states. […] The use of stochastic differential equations for random continuously changing covariates has been studied intensively in the analysis of longitudinal data […] Such a description is convenient since it captures the feedback mechanism typical of biological systems reflecting regular aging-related changes and takes into account the presence of random noise affecting individual trajectories. It also captures the dynamic connections between aging-related changes in health and physiological states, which are important in many applications.”
The links above are links to topics I looked up while reading the second half of the book. The first link is quite relevant to the book’s coverage as a comprehensive longitudinal Grade of Membership (-GoM) model is covered in chapter 17. Relatedly, chapter 18 covers linear latent structure (-LLS) models, and as observed in the book LLS is a generalization of GoM. As should be obvious from the nature of the links some of the stuff included in the second half of the text is highly technical, and I’ll readily admit I was not fully able to understand all the details included in the coverage of chapters 17 and 18 in particular. On account of the technical nature of the coverage in Part 2 I’m not sure I’ll cover the second half of the book in much detail, though I probably shall devote at least one more post to some of those topics, as they were quite interesting even if some of the details were difficult to follow.
I have almost finished the book at this point, and I have already decided to both give the book five stars and include it on my list of favorite books on goodreads; it’s really well written, and it provides consistently highly detailed coverage of very high quality. As I also noted in the first post about the book the authors have given readability aspects some thought, and I am sure most readers would learn quite a bit from this text even if they were to skip some of the more technical chapters. The main body of Part 2 of the book, the subtitle of which is ‘Statistical Modeling of Aging, Health, and Longevity’, is however probably in general not worth the effort of reading unless you have a solid background in statistics.
This post includes some observations and quotes from the last chapters of the book’s Part 1.
“The proportion of older adults in the U.S. population is growing. This raises important questions about the increasing prevalence of aging-related diseases, multimorbidity issues, and disability among the elderly population. […] In 2009, 46.3 million people were covered by Medicare: 38.7 million of them were aged 65 years and older, and 7.6 million were disabled […]. By 2031, when the baby-boomer generation will be completely enrolled, Medicare is expected to reach 77 million individuals […]. Because the Medicare program covers 95 % of the nation’s aged population […], the prediction of future Medicare costs based on these data can be an important source of health care planning.”
“Three essential components (which could be also referred as sub-models) need to be developed to construct a modern model of forecasting of population health and associated medical costs: (i) a model of medical cost projections conditional on each health state in the model, (ii) health state projections, and (iii) a description of the distribution of initial health states of a cohort to be projected […] In making medical cost projections, two major effects should be taken into account: the dynamics of the medical costs during the time periods comprising the date of onset of chronic diseases and the increase of medical costs during the last years of life. In this chapter, we investigate and model the first of these two effects. […] the approach developed in this chapter generalizes the approach known as “life tables with covariates” […], resulting in a new family of forecasting models with covariates such as comorbidity indexes or medical costs. In sum, this chapter develops a model of the relationships between individual cost trajectories following the onset of aging-related chronic diseases. […] The underlying methodological idea is to aggregate the health state information into a single (or several) covariate(s) that can be determinative in predicting the risk of a health event (e.g., disease incidence) and whose dynamics could be represented by the model assumptions. An advantage of such an approach is its substantial reduction of the degrees of freedom compared with existing forecasting models (e.g., the FEM model, Goldman and RAND Corporation 2004). […] We found that the time patterns of medical cost trajectories were similar for all diseases considered and can be described in terms of four components having the meanings of (i) the pre-diagnosis cost associated with initial comorbidity represented by medical expenditures, (ii) the cost peak associated with the onset of each disease, (iii) the decline/reduction in medical expenditures after the disease onset, and (iv) the difference between post- and pre-diagnosis cost levels associated with an acquired comorbidity. The description of the trajectories was formalized by a model which explicitly involves four parameters reflecting these four components.”
As I noted earlier in my coverage of the book, I don’t think the model above fully captures all relevant cost contributions of the diseases included, as the follow-up period was too short to capture all relevant costs to be included in the part iv model component. This is definitely a problem in the context of diabetes. But then again nothing in theory stops people from combining the model above with other models which are better at dealing with the excess costs associated with long-term complications of chronic diseases, and the model results were intriguing even if the model likely underperforms in a few specific disease contexts.
“Models of medical cost projections usually are based on regression models estimated with the majority of independent predictors describing demographic status of the individual, patient’s health state, and level of functional limitations, as well as their interactions […]. If the health states needs to be described by a number of simultaneously manifested diseases, then detailed stratification over the categorized variables or use of multivariate regression models allows for a better description of the health states. However, it can result in an abundance of model parameters to be estimated. One way to overcome these difficulties is to use an approach in which the model components are demographically-based aggregated characteristics that mimic the effects of specific states. The model developed in this chapter is an example of such an approach: the use of a comorbidity index rather than of a set of correlated categorical regressor variables to represent the health state allows for an essential reduction in the degrees of freedom of the problem.”
“Unlike mortality, the onset time of chronic disease is difficult to define with high precision due to the large variety of disease-specific criteria for onset/incident case identification […] there is always some arbitrariness in defining the date of chronic disease onset, and a unified definition of date of onset is necessary for population studies with a long-term follow-up.”
“Individual age trajectories of physiological indices are the product of a complicated interplay among genetic and non-genetic (environmental, behavioral, stochastic) factors that influence the human body during the course of aging. Accordingly, they may differ substantially among individuals in a cohort. Despite this fact, the average age trajectories for the same index follow remarkable regularities. […] some indices tend to change monotonically with age: the level of blood glucose (BG) increases almost monotonically; pulse pressure (PP) increases from age 40 until age 85, then levels off and shows a tendency to decline only at later ages. The age trajectories of other indices are non-monotonic: they tend to increase first and then decline. Body mass index (BMI) increases up to about age 70 and then declines, diastolic blood pressure (DBP) increases until age 55–60 and then declines, systolic blood pressure (SBP) increases until age 75 and then declines, serum cholesterol (SCH) increases until age 50 in males and age 70 in females and then declines, ventricular rate (VR) increases until age 55 in males and age 45 in females and then declines. With small variations, these general patterns are similar in males and females. The shapes of the age-trajectories of the physiological variables also appear to be similar for different genotypes. […] The effects of these physiological indices on mortality risk were studied in Yashin et al. (2006), who found that the effects are gender and age specific. They also found that the dynamic properties of the individual age trajectories of physiological indices may differ dramatically from one individual to the next.”
“An increase in the mortality rate with age is traditionally associated with the process of aging. This influence is mediated by aging-associated changes in thousands of biological and physiological variables, some of which have been measured in aging studies. The fact that the age trajectories of some of these variables differ among individuals with short and long life spans and healthy life spans indicates that dynamic properties of the indices affect life history traits. Our analyses of the FHS data clearly demonstrate that the values of physiological indices at age 40 are significant contributors both to life span and healthy life span […] suggesting that normalizing these variables around age 40 is important for preventing age-associated morbidity and mortality later in life. […] results [also] suggest that keeping physiological indices stable over the years of life could be as important as their normalizing around age 40.”
“The results […] indicate that, in the quest of identifying longevity genes, it may be important to look for candidate genes with pleiotropic effects on more than one dynamic characteristic of the age-trajectory of a physiological variable, such as genes that may influence both the initial value of a trait (intercept) and the rates of its changes over age (slopes). […] Our results indicate that the dynamic characteristics of age-related changes in physiological variables are important predictors of morbidity and mortality risks in aging individuals. […] We showed that the initial value (intercept), the rate of changes (slope), and the variability of a physiological index, in the age interval 40–60 years, significantly influenced both mortality risk and onset of unhealthy life at ages 60+ in our analyses of the Framingham Heart Study data. That is, these dynamic characteristics may serve as good predictors of late life morbidity and mortality risks. The results also suggest that physiological changes taking place in the organism in middle life may affect longevity through promoting or preventing diseases of old age. For non-monotonically changing indices, we found that having a later age at the peak value of the index […], a lower peak value […], a slower rate of decline in the index at older ages […], and less variability in the index over time, can be beneficial for longevity. Also, the dynamic characteristics of the physiological indices were, overall, associated with mortality risk more significantly than with onset of unhealthy life.”
“Decades of studies of candidate genes show that they are not linked to aging-related traits in a straightforward manner […]. Recent genome-wide association studies (GWAS) have reached fundamentally the same conclusion by showing that the traits in late life likely are controlled by a relatively large number of common genetic variants […]. Further, GWAS often show that the detected associations are of tiny effect […] the weak effect of genes on traits in late life can be not only because they confer small risks having small penetrance but because they confer large risks but in a complex fashion […] In this chapter, we consider several examples of complex modes of gene actions, including genetic tradeoffs, antagonistic genetic effects on the same traits at different ages, and variable genetic effects on lifespan. The analyses focus on the APOE common polymorphism. […] The analyses reported in this chapter suggest that the e4 allele can be protective against cancer with a more pronounced role in men. This protective effect is more characteristic of cancers at older ages and it holds in both the parental and offspring generations of the FHS participants. Unlike cancer, the effect of the e4 allele on risks of CVD is more pronounced in women. […] [The] results […] explicitly show that the same allele can change its role on risks of CVD in an antagonistic fashion from detrimental in women with onsets at younger ages to protective in women with onsets at older ages. […] e4 allele carriers have worse survival compared to non-e4 carriers in each cohort. […] Sex stratification shows sexual dimorphism in the effect of the e4 allele on survival […] with the e4 female carriers, particularly, being more exposed to worse survival. […] The results of these analyses provide two important insights into the role of genes in lifespan. First, they provide evidence on the key role of aging-related processes in genetic susceptibility to lifespan. For example, taking into account the specifics of aging-related processes gains 18 % in estimates of the RRs and five orders of magnitude in significance in the same sample of women […] without additional investments in increasing sample sizes and new genotyping. The second is that a detailed study of the role of aging-related processes in estimates of the effects of genes on lifespan (and healthspan) helps in detecting more homogeneous [high risk] sub-samples”.
“The aging of populations in developed countries requires effective strategies to extend healthspan. A promising solution could be to yield insights into the genetic predispositions for endophenotypes, diseases, well-being, and survival. It was thought that genome-wide association studies (GWAS) would be a major breakthrough in this endeavor. Various genetic association studies including GWAS assume that there should be a deterministic (unconditional) genetic component in such complex phenotypes. However, the idea of unconditional contributions of genes to these phenotypes faces serious difficulties which stem from the lack of direct evolutionary selection against or in favor of such phenotypes. In fact, evolutionary constraints imply that genes should be linked to age-related phenotypes in a complex manner through different mechanisms specific for given periods of life. Accordingly, the linkage between genes and these traits should be strongly modulated by age-related processes in a changing environment, i.e., by the individuals’ life course. The inherent sensitivity of genetic mechanisms of complex health traits to the life course will be a key concern as long as genetic discoveries continue to be aimed at improving human health.”
“Despite the common understanding that age is a risk factor of not just one but a large portion of human diseases in late life, each specific disease is typically considered as a stand-alone trait. Independence of diseases was a plausible hypothesis in the era of infectious diseases caused by different strains of microbes. Unlike those diseases, the exact etiology and precursors of diseases in late life are still elusive. It is clear, however, that the origin of these diseases differs from that of infectious diseases and that age-related diseases reflect a complicated interplay among ontogenetic changes, senescence processes, and damages from exposures to environmental hazards. Studies of the determinants of diseases in late life provide insights into a number of risk factors, apart from age, that are common for the development of many health pathologies. The presence of such common risk factors makes chronic diseases and hence risks of their occurrence interdependent. This means that the results of many calculations using the assumption of disease independence should be used with care. Chapter 4 argued that disregarding potential dependence among diseases may seriously bias estimates of potential gains in life expectancy attributable to the control or elimination of a specific disease and that the results of the process of coping with a specific disease will depend on the disease elimination strategy, which may affect mortality risks from other diseases.”
i. “Self-love is often rather arrogant than blind; it does not hide our faults from ourselves, but persuades us that they escape the notice of others.” (Samuel Johnson)
ii. “So much are the modes of excellence settled by time and place, that men may be heard boasting in one street of that which they would anxiously conceal in another.” (-ll-)
iii. “The greatest of faults, I should say, is to be conscious of none.” (Thomas Carlyle)
iv. “Like most of those who study history, he learned from the mistakes of the past how to make new ones.” (Alan John Percivale Taylor)
v. “A method of reasoning may lead to conclusions which are invariably true, even though it start from false premises.” (Francesco Maria Zanotti)
vi. “We are much harder on people who betray us in small ways than on people who betray others in great ones.” (Rochefoucauld)
vii. “Most people are good only so long as they believe others so.” (Friedrich Hebbel)
viii. “Men are more ready to offend one who desires to be beloved than one who wishes to be feared.” (Niccolò Machiavelli)
ix. “Some have been thought brave because they were afraid to run away.” (Thomas Fuller)
x. “Calumny is like counterfeit money: many people who would not coin it circulate it without qualms.” (Diane De Poitiers)
xi. “Let us leave the labels to those who have little else wherewith to cover their nakedness.” (Walter Sickert)
xii. “…there are few truths important enough to justify paining and reproving others for not knowing them…” (Montesquieu)
xiii. “To make astute people believe one is what one is not is, in most cases, harder than actually to become what one wishes to appear.” (Georg Christoph Lichtenberg)
xiv. “It is a trick among the dishonest to offer sacrifices that are not needed, or not possible, to avoid making those that are required.” (Ivan Goncharov)
xv. “It is seldom that the miserable can help regarding their misery as a wrong inflicted by those who are less miserable” (George Eliot)
xvi. “Thousands upon thousands are yearly brought into a state of real poverty by their great anxiety not to be thought poor.” (William Cobbett)
xvii. “The woman whose behavior indicates that she will make a scene if she is told the truth asks to be deceived.” (Elizabeth Jenkins)
xviii. “My business is to teach my aspirations to confirm themselves to facts, not to try and make facts harmonize with my aspirations.” (Thomas Henry Huxley)
xix. “The art of doing mathematics consists in finding that special case which contains all the germs of generality.” (David Hilbert)
xx. “”Obvious” is the most dangerous word in mathematics.” (Eric Temple Bell)
Lately I’ve been reading some of George MacDonald Fraser’s Flashman books, which have been quite enjoyable reads in general; I’m reading the books in the order in which the actions in the books supposedly took place, not in the order in which the books were published, and a large number of the words included below are words I encountered in the first three of the books I read (i.e. Flashman, Royal Flash, and Flashman’s Lady); I decided the post already at that point included a large number of words (the post includes roughly 120 words), so I saw no need to add additional words from the other books in the series in this post as well. I have reviewed a few of the Flashman books I’ve read on goodreads here, here, and here.
i. “No man is rich enough to buy back his past.” (Oscar Wilde)
ii. “There is a luxury in self-reproach. When we blame ourselves we feel that no one else has a right to blame us.” (-ll-)
iii. “Each new generation asks – What is the meaning of life? A more fertile way of putting the question would be – Why does man need a meaning to life?” (Peter Wessel Zapffe)
iv. “One man’s constant is another man’s variable.” (Alan Perlis)
v. “All scientific work is incomplete – whether it be observational or experimental. All scientific work is liable to be upset or modified by advancing knowledge. That does not confer upon us a freedom to ignore the knowledge we already have, or to postpone the action that it appears to demand at a given time.” (Austin Bradford Hill)
vi. “Most women set out to try to change a man, and when they have changed him they do not like him.” (Marlene Dietrich)
vii. “We speak with our lips to explain, with our throats to convince.” (Malcolm de Chazal)
viii. “Those who do not complain are never pitied.” (Jane Austen)
ix. “It is an aspect of all happiness to suppose that we deserve it.” (Joseph Joubert)
x. “Almost all absurdity of conduct arises from the imitation of those whom we cannot resemble.” (Samuel Johnson)
xi. “Nothing so much prevents our being natural as the desire to seem so.” (Rochefoucauld)
xii. “It is harder to hide feelings we have than to feign those we lack.” (-ll-)
xiii. “Almost all our faults are more pardonable than the methods we resort to to hide them.” (-ll-)
xiv. “Fanaticism consists of redoubling your efforts when you have forgotten your aim.” (George Santayana)
xv. “There is nothing that fear or hope does not make men believe.” (Vauvenargues)
xvi. “There is no rule more invariable than that we are paid for our suspicions by finding what we suspected.” (Henry David Thoreau)
xvii. “It is hard to believe that a man is telling the truth when you know that you would lie if you were in his place.” (H.L. Mencken)
xviii. “Conscience is thoroughly well bred and soon leaves off talking to those who do not wish to hear it.” (Samuel Butler)
xix. “History is not written as it was experienced, nor should it be. The inhabitants of the past know better than we do what it was like to live there, but they were not well placed, most of them, to understand what was happening to them and why.” (Tony Judt)
xx. “Stability is much underappreciated, especially by those who enjoy its benefits.” (Curtis Yarvin)
“While the past decade has witnessed a proliferation of neuroimaging studies of autism, theoretical approaches for understanding systems-level brain abnormalities remain poorly developed. We propose a novel anterior insula-based systems-level model for investigating the neural basis of autism, synthesizing recent advances in brain network functional connectivity with converging evidence from neuroimaging studies in autism. The anterior insula is involved in interoceptive, affective and empathic processes, and emerging evidence suggests it is part of a “salience network” integrating external sensory stimuli with internal states. Network analysis indicates that the anterior insula is uniquely positioned as a hub mediating interactions between large-scale networks involved in externally- and internally-oriented cognitive processing. A recent meta-analysis identifies the anterior insula as a consistent locus of hypoactivity in autism. We suggest that dysfunctional anterior insula connectivity plays an important role in autism. […]
Increasing evidence for abnormal brain connectivity in autism comes from studies using functional connectivity measures […] These findings support the hypothesis that under-connectivity between specific brain regions is a characteristic feature of ASD. To date, however, few studies have examined functional connectivity within and between key large-scale canonical brain networks in autism […] The majority of published studies to date have examined connectivity of specific individual brain regions, without a broader theoretically driven systems-level approach.
We propose that a systems-level approach is critical for understanding the neurobiology of autism, and that the anterior insula is a key node in coordinating brain network interactions, due to its unique anatomy, location, function, and connectivity.”
“Participants, 31 recruited via an outpatient clinic and 198 via an online survey, were asked to answer a number of self-report questionnaires. The total sample comprised 229 high-functioning adults with ASD (40% males, average age: 35 years). […] Of the total sample, 73% indicated romantic relationship experience and only 7% had no desire to be in a romantic relationship. ASD individuals whose partner was also on the autism spectrum were significantly more satisfied with their relationship than those with neurotypical partners. Severity of autism, schizoid symptoms, empathy skills, and need for social support were not correlated with relationship status. […] Our findings indicate that the vast majority of high-functioning adults with ASD are interested in romantic relationships.”
Those results are very different from other results in the field – for example: “[a] meta-analysis of follow-up studies examining outcomes of ASD individuals revealed that, [o]n average only 14% of the individuals included in the reviewed studies were married or ha[d] a long-term, intimate relationship (Howlin, 2012)” – and one major reason is that they only include high-functioning autistics. I feel sort of iffy about the validity of the selection method used for procuring the online sample, this may also be a major factor (almost one third of them had a university degree so this is definitely not a random sample of high-functioning autistics; ‘high-functioning’ autistics are not that high-functioning in the general setting. Also, the sex ratio is very skewed as 60% of the participants in the study were female. A sex ratio like that may not sound like a big problem, but it is a major problem because a substantial majority of individuals with mild autism are males. Whereas the sex ratio is almost equal in the context of syndromic ASD, non-syndromic ASD is much more prevalent in males, with sex ratios approaching 1:7 in milder cases (link). These people are definitely looking at the milder cases, which means that a sample which skews female will not be remotely similar to most random samples of such individuals taken in the community setting. And this matters because females do better than males. A discussion can be had about to which extent women are under-diagnosed, but I have not seen data convincing me this is a major problem. It’s important to keep in mind in that context that the autism diagnosis is not based on phenotype alone, but on a phenotype-environment interaction; if you have what might be termed ‘an autistic phenotype’ but you are not suffering any significant ill effects as a result of this because you’re able to compensate relatively well (i.e. you are able to handle ‘the environment’ reasonably well despite the neurological makeup you’ve ended up with), you should not get an autism diagnosis – a diagnostic requirement is ‘clinically significant impairment in functioning’.
Anyway some more related data from the publication:
“Studies that analyze outcomes exclusively for ASD adults without intellectual impairment are rare. […] Engström, Ekström, and Emilsson (2003) recruited previous patients with an ASD diagnosis from four psychiatric clinics in Sweden. They reported that 5 (31%) of 16 adults with ASD had ”some form of relation with a partner.” Hofvander et al. (2009) analyzed data from 122 participants who had been referred to outpatient clinics for autism diagnosis. They found that 19 (16%) of all participants had lived in a long-term relationship.
Renty and Roeyers (2006) […] reported that at the time of the[ir] study 19% of 58 ASD adults had a romantic relationship and 8.6% were married or living with a partner. Cederlund, Hagberg, Billstedt, Gillberg, and Gillberg (2008) conducted a follow-up study of male individuals (aged 16–36 years) who had been diagnosed with Asperger syndrome at least 5 years before. […] at the time of the study, three (4%) [out of 76 male ASD individuals] of them were living in a long-term romantic relationship and 10 (13%) had had romantic relationships in the past.”
A few more data and observations from the study:
“A total of 166 (73%) of the 229 participants endorsed currently being in a romantic relationship or having a history of being in a relationship; 100 (44%) reported current involvement in a romantic relationship; 66 (29%) endorsed that they were currently single but have a history of involvement in a romantic relationship; and 63 (27%) participants did not have any experience with romantic relationships. […] Participants without any romantic relationship experience were significantly more likely to be male […] According to participants’ self-report, one fifth (20%) of the 100 participants who were currently involved in a romantic relationship were with an ASD partner. […] Of the participants who were currently single, 65% said that contact with another person was too exhausting for them, 61% were afraid that they would not be able to fulfil the expectations of a romantic partner, and 57% said that they did not know how they could find and get involved with a partner; and 50% stated that they did not know how a romantic relationship works or how they would be expected to behave in a romantic relationship”
“[P]revious studies that exclusively examined adults with ASD without intellectual impairment reported lower levels of romantic relationship experience than the current study, with numbers varying between 16% and 31% […] The results of our study can be best compared with the results of Hofvander et al. (2009) and Renty and Roeyers (2006): They selected their samples […] using methods that are comparable to ours. Hofvander et al. (2009) found that 16% of their participants have had romantic relationship experience in the past, compared to 29% in our sample; and Renty and Roeyers (2006) report that 28% of their participants were either married or engaged in a romantic relationship at the time of their study, compared to 44% in our study. […] Compared to typically developed individuals the percentage of ASD individuals with a romantic relationship partner is relatively low (Weimann, 2010). In the group aged 27–59 years, 68% of German males live together with a partner, 27% are single, and 5% still live with their parents. In the same age group, 73% of all females live with a partner, 26% live on their own, and 2% still live with their parents.”
“As our results show, it is not the case that male ASD individuals do not feel a need for romantic relationships. In fact, the contrary is true. Single males had a greater desire to be in a romantic relationship than single females, and males were more distressed than females about not being in a romantic relationship.” (…maybe in part because the females who were single were more likely than the males who were single to be single by choice?)
“Our findings showed that being with a partner who also has an ASD diagnosis makes a romantic relationship more satisfying for ASD individuals. None of the participants, who had been with a partner in the past but then separated, had been together with an ASD partner. This might indicate that once a person with ASD has found a partner who is also on the spectrum, a relationship might be very stable and long lasting.”
“The social motivation hypothesis of autism posits that infants with autism do not experience social stimuli as rewarding, thereby leading to a cascade of potentially negative consequences for later development. […] Here we use functional magnetic resonance imaging to examine social and monetary rewarded implicit learning in children with and without autism spectrum disorders (ASD). Sixteen males with ASD and sixteen age- and IQ-matched typically developing (TD) males were scanned while performing two versions of a rewarded implicit learning task. In addition to examining responses to reward, we investigated the neural circuitry supporting rewarded learning and the relationship between these factors and social development. We found diminished neural responses to both social and monetary rewards in ASD, with a pronounced reduction in response to social rewards (SR). […] Moreover, we show a relationship between ventral striatum activity and social reciprocity in TD children. Together, these data support the hypothesis that children with ASD have diminished neural responses to SR, and that this deficit relates to social learning impairments. […] When we examined the general neural response to monetary and social reward events, we discovered that only TD children showed VS [ventral striatum] activity for both reward types, whereas ASD children did not demonstrate a significant response to either monetary or SR. However, significant between-group differences were shown only for SR, suggesting that children with ASD may be specifically impaired on processing SR.”
I’m not quite sure I buy that the methodology captures what it is supposed to capture (“The SR feedback consisted of a picture of a smiling woman with the words “That’s Right!” in green text for correct trials and a picture of the same woman with a sad face along with the words “That’s Wrong” in red text for incorrect trials”) (this is supposed to be the ‘social reward feedback’), but on the other hand: “The chosen reward stimuli, faces and coins, are consistent with those used in previous studies of reward processing” (so either multiple studies are of dubious quality, or this kind of method actually ‘works’ – but I don’t know enough about the field to tell which of the two conclusions apply).
“The idea that social motivation deficits play a central role in Autism Spectrum Disorders (ASD) has recently gained increased interest. This constitutes a shift in autism research, which has traditionally focused more intensely on cognitive impairments, such as Theory of Mind deficits or executive dysfunction, while granting comparatively less attention to motivational factors. This review delineates the concept of social motivation and capitalizes on recent findings in several research areas to provide an integrated picture of social motivation at the behavioral, biological and evolutionary levels. We conclude that ASD can be construed as an extreme case of diminished social motivation and, as such, provides a powerful model to understand humans’ intrinsic drive to seek acceptance and avoid rejection.”
“We examine the nature and predictors of social and romantic functioning in adolescents and adults with ASD. Parental reports were obtained for 25 ASD adolescents and adults (13-36 years), and 38 typical adolescents and adults (13-30 years). The ASD group relied less upon peers and friends for social (OR = 52.16, p < .01) and romantic learning (OR = 38.25, p < .01). Individuals with ASD were more likely to engage in inappropriate courting behaviours (χ2 df = 19 = 3168.74, p < .001) and were more likely to focus their attention upon celebrities, strangers, colleagues, and ex-partners (χ2 df = 5 =2335.40, p < .001), and to pursue their target longer than controls (t = -2.23, df = 18.79, p < .05).”
“Examination of relationships the individuals were reported to have had with the target of their social or romantic interest, indicated that ASD adolescents and adults sought to initiate fewer social and romantic relationships but across a wider variety of people, such as strangers, colleagues, acquaintances, friends, ex-partners, and celebrities. […] typically developing peers […] were more likely to target colleagues, acquaintances, friends, and ex-partners in their relationship attempts, whilst the ASD group targeted these less frequently than expected, and attempted to initiate relationships significantly more frequently than is typical, with strangers and celebrities. […] In attempting to pursue and initiate social and romantic relationships, the ASD group were reported to display a much wider variety of courtship behaviours than the typical group. […] ASD adolescents and adults were more likely to touch the person of interest inappropriately, believe that the target must reciprocate their feelings, show obsessional interest, make inappropriate comments, monitor the person’s activities, follow them, pursue them in a threatening manner, make threats against the person, and threaten self-harm. ASD individuals displayed the majority of the behaviours indiscriminately across all types of targets. […] ASD adolescents and adults were also found […] to persist in their relationship pursuits for significantly longer periods of time than typical adolescents and adults when they received a negative or no response from the person or their family.”