5An Overview of Biomarker Research from Community and Population-Based Studies on Aging

Harris JR, Gruenewald TL, Seeman T.

Publication Details

The goal of this chapter is to provide an overview of findings from community-based studies that have profitably incorporated biomarkers along with more traditional interview data to address important questions regarding factors that affect health risks at older ages. The focus on older age stems from a series of activities (National Research Council, 1997, 2001a) funded by the National Institute on Aging that promoted a new era of population aging research predicated on the importance of integrating biomarkers into survey research.

Prior efforts to include biomarkers in community and population studies have yielded a wealth of knowledge regarding the role of biological systems and processes in cognitive and physical functioning, mental and physical disease development, and mortality outcomes. The biomarker initiative under the National Institute on Aging was particularly concerned with identifying biomarkers of physiological age (Butler et al., 2004). In contrast, biomarker research in the social and behavioral sciences has emphasized elucidating the interplay of biological systems with sociodemographic, behavioral, psychosocial, pharmacological, and genetic factors in health outcomes (Institute of Medicine, 2006; National Research Council, 2001b).

This chapter highlights research contributions from a selected group of community and population studies that include various biomarker measurements in study assessments. It is not intended to review completely, catalogue, or present in-depth results, but rather to provide an overview of the range of biomarker research conducted in these studies and highlight some key findings and research directions stemming from this integrative research.

For the purposes of this review, we have limited our consideration of biomarkers mainly to DNA and physiological biomarkers collected from blood (e.g., glycosylated hemoglobin, cholesterol), saliva (cortisol) or urine (cortisol and catecholamines); not included are more functional parameters that are also considered to be biomarkers, such as hand grip strength, measures of vision or hearing, and assessments of cognitive or physical functioning. This chapter highlights the types of questions that can be addressed when social and behavioral studies are supplemented with biomarker data, and therefore we also exclude descriptions of biomarker research derived from postmortem studies of brain tissue. It is important to note, however, the value of postmortem biomarkers for elucidating biological pathways and mechanisms, including those linking social and behavioral measures with disease outcomes. This is illustrated by research from, for example, the Nun's Study (Snowdon, Kemper, Mortimer, Wekstein, and Markesbery, 1996), the Religious Order Study (Wilson, Bienias, and Evans, 2004), and the Memory and Aging Project (Bennett et al., 2005).

Our decision to focus on physiological parameters and genes was driven in large part by a central goal of the current volume, namely, to help inform social scientists about the potential value of incorporating biomarkers into their projects through exposition of prior research in which analyses of such biomarkers or candidate genes have provided insights into processes and mechanisms affecting healthy aging. In that context, the focus taken herein seems warranted, as such biomarkers have generally not been included in social surveys (whereas assessments of functioning have), so that evidence based on physiological biomarkers is much less well known to the social science community.

Much of the information presented in this chapter derives from the studies summarized in Table 5-1. These were selected to represent a sampling of ongoing community or population-based studies on aging that have collected DNA or other biological or physiological biomarkers and that are not reviewed elsewhere in this volume. We have organized the wide range of findings generated from the studies reviewed according to a number of critical thematic areas that emerged during our review. These include biomarkers and aging, genetic and environmental influences on risk factors for cardiovascular disease (CVD), social and psychological factors, behavior genetics, biomarkers of cognitive aging, biomarkers of physical function and aging, indices of cumulative biological risk, and the relationship between biomarkers and genetic pleiotropy.

TABLE 5-1. Description of Selected Community and Population-Based Studies Conducting Biomarker Research.


Description of Selected Community and Population-Based Studies Conducting Biomarker Research.


Scientists have long been interested in obtaining measures of biomarkers to understand the role of biological systems in functioning and disease processes. Much of this work has been conducted via well-controlled experiments in the laboratory with nonhuman animals or in small-scale human studies focused on specific physiological processes in the lab (e.g., physiological responses to a specific challenge or interaction of specific physiological systems). More recently the value of collecting biomarker measurements in large-scale community and population studies has been recognized (National Research Council, 2001a) and greater emphasis is being placed, by researchers and by funding agencies, on banking biological samples.

The impetus for the collection of biomarkers is to gain a better understanding of the role of specific biological systems in health conditions, including an understanding of the role of biological systems in association with other sociodemographic, behavioral, pharmacological, psychosocial, and genetic contributions to health outcomes. There is a growing literature in the behavioral sciences literature linking social and behavioral factors (ranging from sociocultural and neighborhood influences to interpersonal relations) to biomarkers and health (Berkman and Kawachi, 2000; Cacioppo, Hughes, Waite, Hawkley and Thisted, 2006; Hawkley Masi, Berry, and Cacioppo, 2006; House, Landis, and Umberson, 1988; Kiecolt-Glaser et al., 2005; Ryff and Singer, 2001; Uchino, Cacioppo, and Kiecolt-Glaser, 1996; Wen, Hawkley, and Cacioppo, 2006). These findings, in conjunction with methodological advances, are fostering integrative lines of research to study health and pathways to disease. For example, the recent Institute of Medicine report Genes, Behavior, and the Social Environment (2006) focuses on social environments in the study of gene by environment interactions and health.

A wide variety of biomarkers have been assessed in community- or population-based studies. Biomarkers of cardiovascular, metabolic, endocrine, and immune systems or processes are most commonly assessed. However, other types of biomarker measurements have also been obtained, including exposure indices (e.g., bone and blood lead levels or pesticide blood levels to assess environmental exposure), measurements of vitamin or antioxidant levels, anthropometric measures (e.g., bone length, height, weight), bone density scans, measurements of brain activity (e.g., functional magnetic resonance imaging, fMRI, or electroencephalogram, EEG), as well as markers of the functional status of a bodily system (e.g., forced expiratory volume to assess lung function).

Biomarker values are typically assessed for biological systems at a resting state (e.g., resting blood pressure, blood levels of glucose), but values are also sometimes assessed under conditions of challenge to a system (e.g., blood pressure levels after standing, levels of glucose after a glucose challenge test). Blood, urine, and saliva samples are typical sources for biomarker assessments, although for some of the biomarkers described above, mechanical or electrical devices (e.g., MRI machine) are also used to obtain measurements. The measurement of biomarkers in community- or population-based studies presents a methodological challenge for researchers, as they must determine how and where to obtain biomarker measurements on a large number of people. Most studies have used home-based or clinic-based protocols for the collection of biological specimens, with some investigations utilizing both approaches.

The primary advantage of clinic-based specimen collection is the ability to implement protocols requiring greater temperature control (e.g., samples must be kept cold or iced) as well as meeting more restricted processing requirements (e.g., within minutes or several hours at most). However, sample representativeness can sometimes suffer as certain subgroups are less able or willing to come to a clinic for reasons related to health, transportation, or unfamiliarity with the location. Examples of studies that have used clinic-based protocols successfully include the Women's Health and Aging Studies I and II, the Cardiovascular Health Study, and the Health, Aging and Body Composition Study. Studies using home-based protocols include the MacArthur Study of Successful Aging (Berkman et al., 1993), the Later Life Resilience Study (Ryff, Singer, and Dienberg Love, 2004), and the Swedish Twin Studies (Lichtenstein et al., 2002).


Most established biomarkers indices reflect age norms in disease-free samples from which individuals with known risks and diseases are excluded from study. This poses a challenge for aging studies because these criteria make it difficult to define “normal” values among groups of elderly individuals for whom morbidity is common. Furthermore, this approach could mask age changes in many biomarkers and values of routine biochemical blood tests. A study of clinical biochemical values in a population-based sample of twins ages 82 and older from the Swedish study of Origins of Variance in the Oldest-Old: Octogenarian Twins (OCTO-Twin) found few participants without clinical diagnoses; therefore, subsequent survival for six years was used as a marker of overall health in late life. Results revealed an association between mortality and higher serum levels of urea, urate, gamma-glutamyltransferase (gamma-GT), free thyroxin, and plasma homocysteine. In women, increased mortality was associated with low serum values for albumin and total cholesterol. The authors propose that these results could provide guidelines for clinical practice and general health examinations (Nilsson et al., 2003a, 2003b).

Further study of the association of biochemical values with morbidity, drug therapy, and anthropometry was examined. In addition to expected findings showing that biochemical values deviate under disease states common among the elderly, a number of biological risk factors exhibit patterns of increasing risk with age, including blood pressure, glucose, and markers of inflammation and homocysteine, each of which is associated with risks for one or more common diseases of aging, such as cardiovascular disease, osteoporosis, hip fracture, depression, and dementia (Nilsson et al., 2003a, 2003b). These findings indicate that morbidity and health-related factors common in aging populations substantially influence routine biochemical values.

Analyses of community-based studies, such as the MacArthur Study of Successful Aging, also point to potentially important age-related reductions in risks associated with some biological factors (e.g., a reduction in the apparent risks associated with elevated total cholesterol—Karlamangla, Singer, Reuben, and Seeman, 2004), although other major risk factors continue to exhibit strong effects with respect to risks for major outcomes, such as physical function (Reuben et al., 2002), cognitive function (Weaver et al., 2002), and longevity (Hu et al., 2005). Analyses of biomarker data from the Cardiovascular Health Study (CHS) have confirmed that lipid profiles (specifically total and low-density lipoprotein [LDL], cholesterol) are not significant predictors of myocardial infarction (MI), stroke, or mortality among older adults (Psaty et al., 2004).

CHS data also point to the continued importance of such biomarkers as high blood pressure, fasted glucose, low albumin, elevated creatinine, and low forced vital capacity as significant, independent risk factors for mortality, along with additional measures of subclinical disease, including aortic stenosis, abnormal left ventricular ejection fraction, major electrocardiographic abnormalities, and stenosis of the internal carotid artery (Fried et al., 1998) and markers of inflammation (Jenny et al., 2006). Analyses from the Health, Aging and Body Composition Study (Health ABC) (another cohort study of adults ages 70 and older) suggest that the presence of metabolic syndrome (based on a complex of risk factors, including cholesterol, blood pressure, and glucose) in older adults does continue to predict subsequent coronary events, heart failure, myocardial infarctions, and cardiovascular-related mortality (Butler et al., 2006). Like the MacArthur and other studies of aging, Health ABC data indicate that inflammatory markers such as interleukin-6 (IL-6) and tumor necrosis factor (TNF)-α are associated cross-sectionally with the presence of subclinical or clinical cardiovascular disease (Cesari et al., 2003a), and the presence of peripheral artery disease (McDermott et al., 2005) and that high levels of IL-6, TNF-α, and C-reactive protein (CRP) also prospectively predict incident coronary events, including coronary heart disease, stroke, and congestive heart failure (Cesari et al., 2003b).

A recent analysis of time trends in biological markers of health risks in participants ages 65 and older from the National Health and Nutrition Examination Surveys (NHANES) III (1988-1994) and NHANES IV (1999-2000) found significant reductions in the levels of high-risk indices of total cholesterol and homocysteine. These findings are consistent with the hypothesis that lipid-lowering medications and folate supplementation, respectively, have been effective in reducing the prevalence of “high-risk” values for these biological parameters. However, other changes indicate an increased burden associated with obesity-related measures and high-risk levels of CRP (Crimmins et al., 2005).

Analyses based on available biomarker data in the Women's Health and Aging Studies (WHAS I and II) have yielded insights into relationships of novel and potentially modifiable biomarkers, such as serum selenium and carotenoids and mortality—low levels in each case predicting higher mortality risk (Ray et al., 2006). Analyses also confirmed mortality risks associated with gradations of diabetic hyperglycemia (Blaum et al., 2005a).

Another example of age effects is illustrated by studies of mortality and the apolipoprotein E (ApoE) genotype. The ApoE gene is a cholesterol transporter involved in brain repair mechanism and is among the most widely studied genes in aging research. It has three common alleles, ApoE-2, ApoE-3, and ApoE-4, with ApoE-4 conferring a higher risk for developing Alzheimer disease (AD). Demographic models investigating the association between ApoE genotype and mortality rates using data from several countries revealed diminishing risks with age and that the ApoE genotypes are associated with little variation in mortality among centenarians (Ewbank, 2002).


In certain respects, the strategy behind biomarker research reflects an endophenotype (intermediary phenotype) approach, described in psychiatric genetics (Gottesman and Gould, 2003). In polygenic systems, in which environmental influences also figure prominently, biomarkers often represent “downstream” traits in the pathways between genes and a measured behavior or outcome. Endophenotype-based strategies are becoming more common in gene detection studies, and it is critically important to recognize that many biomarkers are themselves complexly determined. As illustrated by studies of CVD risk factors, genetic and environmental analyses of biomarkers can provide insights regarding biomarker effects in complex causal pathways.

CVD risk factors are among the most studied biomarkers. They are well known, easy to measure, and routinely assayed in clinical practice. Heritable and environmental factors have long been known to play an important role in cardiovascular disease, although quantification of these risks in the population at large is not straightforward. Twin and family studies provide ample evidence that biological risk factors (biomarkers) for cardiovascular disease, such as serum lipids and apolipoproteins, are also influenced by multiple genetic and environmental factors. Developmental trends in lipid and apolioproteins have been described (e.g., early life, adolescence, menopause in women, and older age) and age-sex differences emerge in lipid profiles.

A key question is how variation in these age and sex patterns is differentially influenced by genes and environment. Research on CVD risk factors from the Swedish Adoption Twin Study of Aging (SATSA) have helped to elucidate this. Genetic differences account for more than half the variation in plasma factor VII levels (Hong, Pedersen, Egberg, and deFaire, 1999). Analyses of total cholesterol, high-density lipoprotein (HDL), apolipoproteins A-I and B, and triglycerides revealed substantial heritabilities for these measures ranging from 0.63 to 0.78 in the younger group (ages 52-65) and from 0.28 to 0.55 in the older group (ages 66-86) for these measures. Heritabilities were consistently lower in the older age groups and significantly lower for apolipoprotein B and triglyceride levels (Heller, de Faire, Pedersen, Dahlen, and McClearn, 1993).

Furthermore, the unique design involving twins reared apart and reared together revealed that shared rearing environmental effects contributed to variation in total cholesterol levels. Most of the nongenetic variation for the other biomarkers was explained by unique environmental factors. This research was expanded on in a further twin study using augmented data that included SATSA, the Screening Across the Lifespan Twin (SALT) pilot study, and the Men and Women's Aging (GENDER) study. Sex and age differences in lipid and apolipoprotein levels were investigated in three age groups (17-49, 50-69, and 70-85). Heritabilities ranged from 35 to 74 percent and were consistently greater in women across all age groups. This age effect is almost entirely explained by greater variance associated with environmental factors (Iliadou, Lichtenstein, deFaire, and Pedersen, 2001). Age dependency in genetic effects for plasma lipids was also investigated in Dutch data, which showed some evidence for partially different sets of genes to influence lipid levels at different ages (Sneider, Van Doornen, and Boomsma, 1997).

Numerous other studies have confirmed the importance of genetic and environmental factors on cardiovascular risk factors in different countries (see Iliadou and Sneider, 2004, for a review), illustrating that genetic influences differentially affect various risk factors. For example, heritabilities for lipoprotein a Lp(a) levels are exceptionally high (.90) but may be affected by the use of sex hormones among women (Hong et al., 1995). Future research that incorporates measured environments into these models will help identify the nature of nonshared influences that impinge on CVD risk factors and become more important with age and, potentially, impact CVD outcomes.


A large and rapidly growing literature documents links among social, behavioral, and psychosocial measures with physiological, biological, and genetic factors (Harris, 2007; Ryff and Singer, 2005). A key goal of many of the studies reviewed here is to understand the nature of some of these links and to elucidate the interplay between these factors and their relationships to biology as these latter relationships affect healthy aging. This research has investigated an array of factors spanning different spheres of influence, including socioeconomic status, social engagement or social relationships, and various psychological factors.

Socioeconomic Status

Although previous research has suggested that socioeconomic status (SES) is less strongly related to health outcomes at older ages (House et al., 1994), data from multiple population-based studies point to the continued impact of lower SES on health risks in later life, including links to greater biological dysregulation in various major regulatory systems. Multiple studies based on data from the Health ABC Study have documented relationships between indicators of SES and biomarkers and also provide evidence that lower SES confers a greater risk for poorer cognitive and physical functioning outcomes. Lower SES (assessed by levels of education, income, and assets) was associated with higher levels of CRP, TNF-α, and IL-6 (Koster et al., 2006). Behavioral factors, such as smoking, alcohol consumption, and obesity accounted for a greater proportion of these associations than prevalent disease (e.g., heart disease, diabetes).

Lower SES participants were also at greater hazard of developing mobility disability (Koster et al., 2005a). Body mass index and an index of the number of inflammatory biomarkers (CRP, IL-6, TNF-α) for which participants' values were in the highest tertile were significant covariates in analyses, suggesting that these biomarkers may be important mediators of relationships between SES and mobility disability.

Lower SES has also been shown to predict the likelihood of cognitive decline in Health ABC participants (Koster et al., 2005b). Although a number of biomarkers were associated with both SES and cognitive decline in this investigation, biomarkers explained only a small portion of the relationship between SES and cognitive decline.

Data from the MacArthur Study of Successful Aging document similar SES gradients in major biomarkers (lower SES being associated with increased prevalence of higher risk levels), with a significant, negative gradient seen for a summary index of allostatic load (Seeman et al., 2004a). Analyses also indicate that these SES differences in allostatic load mediate 35 percent of the SES gradient in mortality (Seeman et al., 2004a). Parallel evidence for such SES gradients in cumulative biological risk has been reported from the Wisconsin Longitudinal Study (Singer and Ryff, 1999). And low levels of educational attainment were also shown to be associated with higher levels of allostatic load in the Normative Aging Study (NAS), with much of the association apparently mediated by higher levels of hostility (Kubzansky, Kawachi, and Sparrow, 1999). Recent analyses of NHANES III data also show consistent SES gradients in levels of biological risks with respect to blood pressure, cholesterol, metabolic profiles, and levels of inflammation (Seeman, Merkin, Crimmins, Koretz, and Karlamangla, no date).

Social Engagement and Social Relationships

A consistent body of evidence links greater social integration to better health and longevity (Seeman, 1996). A growing literature further indicates that health effects of social relationships persist importantly into older age, with evidence linking greater social integration or engagement to lower risks for cognitive decline (Bassuk, Glass, and Berkman, 1999; Seeman, Lusignolo, Albert, and Berkman, 2001a) and dementia (Fratiglioni, Wang, Ericsson, Maytan, and Winblad, 2000), as well as lower risks for physical disability (Seeman, Bruce, and McAvay, 1996) and greater longevity (Seeman, Kaplan, Knudsen, Cohen, and Guralnik, 1987; Seeman et al., 1993).

More recently, attention has shifted to the identification of potential biological pathways through which social relationships affect health and aging. Greater social integration and higher levels of reported emotional support have been linked to lower levels of major stress hormones (e.g., cortisol, norepinephrine, epinephrine) among older men, with weaker and nonsignificant trends among women (Seeman, Berkman, Blazer, and Rowe, 1994), and similar findings have been reported more recently for patterns of association with IL-6 (Loucks, Berkman, Gruenewald, and Seeman, 2006). Examination of more cumulative measures of biological risk has yielded evidence of an inverse relationship between levels of positive social engagement and cumulative risk for both men and women (Seeman, Singer, Ryff, and Levy-Storms, 2002). Importantly, data on levels of social conflict point to the increased health risks associated with greater exposure to such interactions (see Burg and Seeman, 1994, for a review).

Paralleling these earlier data, more recent data provide a growing body of evidence indicating that such negative social interactions are associated with heightened physiological activity and reactivity, resulting in significantly increased levels of biological risk (Seeman and McEwen, 1996; Taylor, Repetti, and Seeman, 1997; Seeman et al., 2002). These findings are consistent with the fact that man has evolved as a social animal. As such, it is not surprising that qualitative aspects of our social interactions should be associated with underlying and parallel patterns of physiological activation reflecting reduced versus heightened biological risk profiles.

Psychological Factors

The Later Life Resilience Study is another investigation that has focused on links between biomarkers and a number of social and psychological factors. Using a sample of women drawn from a larger study of the process of relocating to “senior housing,” Ryff and colleagues (2006) examined biomarker correlates of positive and negative well-being in study participants to address the question of whether positive and negative well-being are opposite sides of a single mental health continuum or whether these forms of well-being are separate and independent dimensions. They hypothesized that if positive and negative well-being are unique constructs, they should have distinct biomarker correlates, and this premise was largely supported by study findings.

For example, salivary cortisol, norepinephrine, waist-hip ratio, HDL cholesterol, and total/HDL cholesterol showed significant associations with some measures of well-being (e.g., purpose in life, personal growth, perceived autonomy, positive affect, positive relations) but not with measures of ill-being, while dehydroepiandrosterone sulfate (DHEAS) and systolic blood pressure showed significant associations with negative well-being (e.g., depressive symptoms, negative affect, trait anxiety, trait anger) but not positive well-being. Only two biomarkers, weight and glycosylated hemoglobin, showed significant associations with indicators of both positive and negative well-being.

Other analyses have examined links between biomarkers and two dominant forms of positive well-being, eudaimonic (e.g., self-actualization, purpose in life) and hedonic (e.g., happiness, pleasure, satisfaction) well-being. A number of biomarkers, including salivary cortisol, urinary norepinephrine, the soluble receptor for IL-6, weight, waist-hip ratio, HDL cholesterol, total/HDL cholesterol, glycosylated hemoglobin, and indicators of better sleep quality were found to relate to scores on measures of eudaimonic well-being, while only HDL cholesterol was associated with hedonic well-being (Ryff et al., 2004).

Women with more positive social relations have also been found to have lower levels of IL-6 and better sleep efficiency (Friedman et al., 2005). Interestingly, social relations and sleep quality showed an interaction in their association with IL-6, such that those with both poor social relations and sleep efficiency had the highest levels of IL-6, while those women with poor scores on only one measure showed more moderate levels. This finding suggests that social relationships and sleep may act to buffer one another for deficits in one domain in terms of potential impacts on inflammatory activity. Evidence that poor sleep may be related to CVD risk, particularly in older people under chronic stress, has also been reported. Poor sleep, as measured using polysomnography, was associated with higher plasma IL-6 and procoagulant marker fibrin D-dimer in a study of older community-dwelling caregiver and noncaregiver adults, and these effects were more pronounced in caregivers of patients with Alzheimer disease (von Kanel et al., 2006).

Data from the Normative Aging Study also point to the potential importance of hostility as a factor affecting patterns of biological risk as well as major health outcomes. For example, high hostility has been linked to high levels of insulin and triglycerides and low HDL levels, with much of these associations mediated by links between higher hostility and higher body mass index and waist-hip ratio (Niaura et al., 2000). As indicated earlier, high levels of hostility have also been found to be associated with high scores on an allostatic load index, which represented a summary score of high-risk values on a range of cardiovascular, endocrine, and metabolic biomarkers (Kubzansky, Kawachi, and Sparrow, 1999). Investigations of NAS participants have also examined the interaction of both biological and psychosocial risk factors in the development of disease. For example, one investigation found that high levels of hostility and the presence of metabolic syndrome (high-risk scores on measures of triglycerides, glucose, HDL cholesterol, blood pressure, and body mass index [BMI]) independently predicted the likelihood of myocardial infarction over a 14-year follow-up period, but the concomitant presence of both high hostility and metabolic syndrome was associated with the greatest risk for subsequent myocardial infarction (Todaro et al., 2005).

Chronic stress has also been linked to a myriad of effects, including poor health, reduced immune function, premature aging, less robust responses to treatment, and earlier age of disease onset. Epel and colleagues (2004) analyzed peripheral blood mononuclear cells in premenopausal women to investigate the relationship between life stressors (chronic caretaking of an ill child) and cellular aging. Although there was no relationship between caregiver status and the indices of cellular aging, results revealed that perceived stress and duration of stress were associated with higher oxidative stress, lower telomerase activity, and shorter telomere length (TL), even after controlling for age. This research provides insight into potential mechanisms that link psychological stress to biological aging and suggests that perceived or chronic stress may play a role in premature senescence. Longitudinal research that tracks changes in telomere length in association with other psychosocial and health indices is needed to provide further insights into the factors affecting telomere shortening across the lifespan. Although this research is not based upon a community sample, measures of telomere length and telomerase activity are becoming more common biomarkers in larger cohort studies. For example, in the MacArthur Study of Successful Aging, stored DNA has been used to assess telomere length, and analyses now under way are examining both relationships to subsequent health risks and predictors of TL.


Following the human genome project there has been a tremendous shift in the focus of behavioral genetic research that has moved this field more squarely into the realm of biomarker research. A large body of research unequivocally documents the importance of anonymous genetic influences for an array of behaviors that affect healthy aging, including symptoms of depression, physical functioning, personality, health-related behaviors, and cognition (see special issue of Behavior Genetics on Aging, 33(2), March 2003).

Candidate genes, and their protein products, represent biomarkers that are becoming more routinely catalogued and studied. For example, based on information regarding relevant biological pathways, projects from the Swedish twins (SATSA, OCTO-Twin, GENDER, HARMONY, and SALT) have identified a variety of candidate genes and gene markers to study in association with aging-related behaviors and health outcomes. An example finding from this approach and combining data from several studies (SATSA, OCTO-Twin, and GENDER) is the reported association between depressed mood in the elderly and the AA gene variant in the serotonin receptor gene (5-HTR2A) among men but not among women (Jansson et al., 2003). These results raise the question whether different genes or genetic mechanisms contribute to the development of depressed mood in the elderly.

Methodological advancements have added a new focus to behavior genetic research, and an increasing number of studies are interested in the localization and identification of functional genetic variants influencing individual differences in human behavior. For example, loneliness is common among the elderly and is central to a cluster of socioemotional states and affects behavior, psychological health (Cacioppo et al., 2006), and physical health (Tomaka, Thompson, and Palacios, 2006). Social isolation and loneliness are implicated in the pathogenesis of multiple diseases, responses to therapy, and mortality (Hawkley and Cacioppo, 2003). Putative disease pathways through which loneliness exerts an influence include health behaviors, excessive stress reactivity, and deficiencies in physiological repair and maintenance. Research indicates that genetic differences among people explain about half of the variation in loneliness among children (McGuire and Clifford, 2000) and adults (Boomsma, Willemsen, Dolan, Hawkley, and Cacioppo, 2005).

Data from the Netherlands Twin Registry (Boomsma et al., 2002) were analyzed to investigate the molecular-genetic basis for these findings. Genotypic marker (400 microsatellite markers) data were collected in 682 sibling pairs and their parents. Linkage and association analyses were conducted to elucidate candidate regions that may contain genes that influence variation in loneliness. Results pointed to a region on chromosome 12q23-24, and follow-up association tests showed significant association to two neighboring markers, D12S79 and D12S395. Linkage results in this region have been reported previously for a number of psychiatric disorders and neuroticism. Although the collective linkage results are not definitive or consistent, they provide enough evidence for the authors to postulate that this region on chromosome 12 may contain genes involved in affective and social regulation and dysregulation (Boomsma, Cacioppo, Slagbom, and Posthuma, 2006).

Although linkage studies per se do not constitute strict biomarker research, the effort to identify biomarkers of complex behaviors is becoming more and more common, as illustrated by a recent special issue of Behavior Genetics, 36(1), January 2006, dedicated to genetic linkage studies for behavioral traits, including emotionality, depression, loneliness, cognition, addictive behaviors, health behaviors, and their endophenotypes. Such an approach may potentially generate numerous useful biomarkers that may become standard covariates in large cohort studies. Aided by publicly available databases, the value of these data for exploring complexly determined phenotypes could be greatly enhanced through cross-study comparisons and data pooling possibilities.

Recently, attention has focused on the importance of studying gene-environment interactions, including how behaviors may modify gene expression (Harris, 2007; Rutter, Moffit, and Caspi, 2006). One of the most extensively studied common genetic variants in human studies of social behavior is the 5′-promoter polymorphism of the serotonin transporter gene (5HTT). A wave of studies has explored interactions between functional polymorphisms of this gene with stressful life events and depression. Evidence supporting such an interaction effect is reported by some (Caspi et al., 2003; Eley et al., 2004; Grabe et al., 2005; Kaufman et al., 2004; Kendler, Kuhn, Vittum, Prescott, and Riley, 2005) but not confirmed by all (Gillespie, Whitfield, Williams, Heath, and Martin, 2005; Surtees et al., 2006) studies. Although these findings do not derive from aging studies per se, they generate important aging-related questions regarding the role of functional polymorphisms in the serotonin transporter gene and late life depression.

New findings from the Swedish SATSA study report evidence for genotype by environment interactions affecting change in a semantic memory task. This work investigated the role of several candidate genes including those coding for ApoE and estrogen receptor alpha (ESR1) and serotonin candidates (HTR2A and 5HTT). Further investigation aimed at identifying the nature of the environmental influences involved in this interaction examined social and stress factors, including social support, life events, and depressive symptoms. Results suggested that influences associated with depressive symptoms may moderate the gene-environment interaction observed for ESR1 and ApoE and longitudinal semantic memory change. The authors explain that noncarriers of putative risk alleles may be relatively more sensitive to depression-evoking environmental contexts than carriers of the risk allele. This suggests that the contexts that facilitate or reduce depressive symptoms could affect resiliency in semantic memory dependent on genotype (Reynolds, Gatz, Berg, and Pedersen, 2007).


Cognition represents one of the two major domains of functioning, the other being physical functioning (see below). Among the biomarkers known to affect risks for cognitive decline, the ApoE genotype, more specifically the ApoE-4 allele, has received by far the most attention to date. After the initial discovery of the role of ApoE-4 for dementia, numerous works have examined the role of ApoE and the ApoE-4 variant in trajectories of normal aging as well as in the pathogenesis of other diseases. For example, analysis of stored samples in the MacArthur study confirmed that ApoE status is predictive of cognitive declines in this initially high-functioning cohort (Bretsky, Guralnik, Launer, Albert, and Seeman, 2003). Studies of diverse populations reveal that the risk conferred by the ApoE-4 allele does not pertain to all populations. Results from a study of healthy Medicare recipients in the Washington Heights–Inwood community of New York City (WHICAP) revealed that ApoE-4 conferred a greater risk for Alzheimer disease among non-Hispanic whites, but that blacks and Hispanics were at increased risk regardless of their ApoE genotype (Tang et al., 1998). These results indicate that other, unknown genetic or environmental risk factors contribute to the increased risk of Alzheimer disease in blacks and Hispanics.

A particularly exciting opportunity is offered by the growing availability of more population-based genetic information in studies that include a rich array of other information. These data offer extraordinary new possibilities to investigate the ways in which risks associated with genotype may be importantly modified by other characteristics (i.e., to test for interactions). Such research, particularly as it relates to potentially modifiable characteristics of individuals (or environments), may offer important insights regarding protective factors that can compensate for and reduce ultimate risks associated with particular genotypes—factors that could be the focus of interventions to reduce health risks. One such question has been whether acquisition of higher levels of education might provide a protective effect against the known risks for cognitive decline usually associated with the ApoE-4 allele. Results to date are mixed, although comparisons across studies are hampered by differences in study populations and, perhaps more importantly, by lack of comparability in measures used to assess cognitive aging. Findings are nonetheless illustrative of the potential for use of growing genotypic data in the context of population studies with information on other characteristics of the individual, the environment, or both. For example, analysis from the MacArthur Successful Aging Study revealed that the presence of the ApoE-4 allele was associated with greater declines in cognitive performance (based on detailed assessments of major domains of cognitive function, including naming, spatial recognition, praxis, and executive function) over a seven-year follow-up among the more educated but not among those with less than high school education; risks for cognitive decline were highest and comparable for those with and without the ApoE-4 allele (Seeman et al., 2005).

Another epidemiological study from Washington state has examined this question using a community-dwelling (n = 2,168) sample of nondemented elderly who were followed prospectively for six years using the cognitive abilities screening instrument. With their larger sample, analyses of gene-dose effect were possible and provided evidence for biological effects of the ApoE-4/ApoE-4 genotype compared with the heterozygous and other homozygous configurations. Analyses of a possible interaction with education yielded evidence for education modification of effects only among those with the ApoE-4/ApoE-4 genotype; cognitive decline was greater among those with less education (Shadlen et al., 2005). Differences in analytic approach and operationalization of ApoE status preclude direct comparisons to the MacArthur findings. However, the findings do appear to contrast with trends seen in the MacArthur study, in which greater overall declines were associated with the presence of the ApoE-4 allele among those with higher education (though they remained at higher levels of function than carriers of the ApoE-4 allele who had lower education throughout the follow-up). However, comparisons between these studies are hampered by noncomparability on outcome measures as well as study populations. Nonetheless, their findings suggest that further attention to the joint effects of ApoE and education with respect to cognitive aging are clearly merited.

Along similar lines to the foregoing studies are efforts to investigate the potential of early life socioeconomic environment to modify the relationship between ApoE status and risk for development of Alzheimer disease. Census data were used to index socioeconomic risk based on a number of parental and demographic measures. In addition to the increased risk conferred by genetic predisposition and early life environment, risk for AD was found to be greatly elevated (OR = 14.8; 95% CI, 4.9-46) when both the genetic and the environmental risk factors were present (Moceri et al., 2001). Other data from the CHS also point to possibly important interactions with dietary fatty fish consumption in relation to dementia risk: dietary intake had a significant effect only for those without the ApoE-4 allele (Huang et al., 2005).

Findings from community-based studies on the effect of ApoE on memory performance and memory change are mixed, with some finding deficits in performance or quicker rates of decline and others reporting no effect. However, ApoE-4 influences are more consistently reported for episodic versus working memory.

Two other genes with mixed support as genetic risk factors for Alzheimer disease, A2M (alpha-2-macroglobulin) and low-density LRP (lipoprotein receptor-related protein), have now been studied in relation to memory among nondemented adults. Variation in these three genes was analyzed in latent growth models measuring memory performance over a 13-year period in SATSA. Polymorphisms of ApoE and A2M (but not low-density LRP) were associated with memory performance and change in memory in this nondemented sample. Specifically, ApoE status affected ability levels of working and recall memory, the ApoE-4/ApoE-4 genotype was associated with worst performance across all ages. Furthermore, this study provided evidence of within-locus interactions (genetic dominance deviations) because memory performance among the heterozygotes was better than among the noncarriers of the ApoE-4 allele. Finally, the rare del/del genotype of A2M was associated with a more rapid rate of decline on figural recognition than the other two genotypes (Reynolds et al., 2006).

The role of other candidate genes has recently been explored in relation to cognitive aging. Age-related loss of serotonin receptors 2A (5-HT2A) is associated with a loss in brain regions including the hippocampus and posterior medial prefrontal cortex. A functional variant (H452Y) of the gene coding for the 5-HT2A serotonin receptor (HTR2A) has been associated with recall tests in young adults (de Quervain et al., 2003). This was further investigated using a new approach that explored allelic associations and trajectories of change in memory performance over a 13-year period in the SATSA data. Findings suggested that the 5-HT2A serotonin receptor is involved in the formation of episodic memories in older adults. Performance on figural memory at age 65 and change in figural memory were associated with the HTR2A genotype. Genotype-dependent effects comparing the AG, AA, and AG configurations revealed the steepest declines associated with AG heterozygotes. Performance over time was consistently worse among those with the AA compared with the GG genotypes, with trajectories differing by 2-6 percent per year (Reynolds, Jansson, Gatz, and Pedersen, 2006).

Other biomarkers also show an association with cognitive decline. Data from the OCTO-Twin study revealed lower homocysteine values among those with intact cognitive function, a finding that contrasts results regarding dementia. These findings overlap somewhat with those from the MacArthur Study of Successful Aging and the NAS, both of which found that high levels of homocysteine and low levels of vitamin B and folate were associated with cognitive decline (Kado et al., 2005; Tucker, Qiao, Scott, Rosenberg, and Spiro, 2005). Elevations in various stress hormones (i.e., urinary free cortisol and/or epinephrine) have also been shown to predict increased risks for cognitive decline (Seeman, McEwen, Singer, Albert, and Rowe, 1997a; Karlamangla, Singer, Chodosh, McEwen, and Seeman, 2005b; Karlamangla, Singer, Greendale, and Seeman, 2005a). Additional biomarkers found to predict cognitive and mental health status, independent of other standard sociodemographic and lifestyle risk factors, include low serum thyroxine (a marker of thyroid function) as a risk factor for cognitive decline (Volpato et al., 2002) and vitamin B12 deficiency as a risk factor for depression (Penninx et al., 2000). High levels of lead in bone and blood have also been shown to be associated with cognitive impairment (Payton, Riggs, Spiro, Weiss, and Hu, 1998). Greater vitamin E intake or vitamin E levels have also been associated with less cognitive impairment and dementia in the InChianti study (Cherubini et al., 2005).

More detailed neuropsychological protocols have also provided evidence linking early cognitive test performance to risks for Alzheimer disease (Saxton et al., 2004), and more detailed MRI data show that levels of inflammation (e.g., fibrinogen) and forced vital capacity are positively and significantly related to white matter disease (Ding et al., 2003).


Physical functioning represents the second major functional domain and has been a focus of considerable research attention. Among the most detailed of this work has been that by Fried and colleagues as part of the WHAS. A particular strength of the WHAS I and II studies is their detailed evaluation of both lower and upper extremity function, including various assessments of balance, walking speed, timed chair stands and knee extension force for lower extremity function and hand grip, pegboard, and putting-on-blouse tests for upper extremity function. Analyses based on these data have documented the significant contribution of muscle weakness to risks for disability (Rantanen et al., 1999) and the more general contributions of physical performance across the various domains of upper and lower extremity functioning to both progressive and catastrophic disability (Onder et al., 2005). In the WHAS I and II, availability of detailed performance measures, plus information on biological processes thought to impact such performance, such as inflammation or growth factors, has afforded important opportunities to examine the joint and independent contributions of performance abilities and biological processes to actual levels of functional disability.

Several analyses have highlighted the significant, negative impact of higher burdens of inflammation in terms of reduced muscle strength and declines in physical function (Ferrucci et al., 2002) as well as parallel negative associations between low insulin-like growth factor-1 (IGF-1) and declining IGF-1 with slower walking speed and reported difficulty with mobility tasks, respectively (Cappola, Bandeen-Roche, Wand, Volpato, and Fried, 2001). These findings on the relationship between muscle strength and inflammation mirror results from Health ABC showing that high levels of oxidized LDL (oxLDL) and IL-6 predict incident mobility disability, with those individuals with high levels of both biomarkers being at greatest risk. Loss of muscle strength and muscle mass may be one pathway involved in relationships between inflammation and physical disability, as high circulating levels of inflammatory biomarkers are associated with low muscle strength and muscle mass in study participants (Visser et al., 2002; Yende et al., 2006).

Inflammatory markers have also been linked to both levels of physical activity and risks for functional decline. Data from the InChianti (Invecchiare in Chianti, meaning “aging in the Chianti area”) Study, a prospective investigation of over 1,000 community-dwelling French adults, indicate that compared with sedentary individuals, physically active men have lower fibrinogen, CRP, IL-6, and TNF-α, as well as lower uric acid and a lower erythrocyte sedimentation rate, while physically active women have lower CRP, IL-6, and uric acid (Cherubini et al., 2005; Elosua et al., 2005). Similar findings have been reported from the MacArthur Study of Successful Aging in the United States (Reuben, Judd-Hamilton, Harris, and Seeman, 2003). Greater inflammatory burden has also been associated with increased risks for physical disability (Reuben et al., 2002) and with poorer physical performance (Cesari et al., 2005).

A number of investigations in InChianti have also examined biomarker correlates of anemia and links between anemia and indicators of physical health or functioning. Anemia is typically defined by low levels of blood hemoglobin (< 12 g/dL in women, < 13 g/dL in men). Hemoglobin is found in red blood cells, which contain iron and are responsible for carrying oxygen to bodily tissues. Anemia or low blood hemoglobin levels are associated with low muscle density, low skeletal muscle strength, a low muscle/total area ratio, and low bone mass and density (Cesari et al., 2005). Consistent with these bone and muscle correlates of anemia, anemic persons have also been found to be more likely to have physical disabilities, poor physical performance, lower hand grip strength, and lower knee extensor strength, compared with nonanemic individuals (Penninx et al., 2005).

Data from the InChianti study also point to the importance of additional biomarkers, including antioxidants such as vitamin E (alpha-tocopherol), which has been associated with a number of indicators of health and functioning in study participants. Those with lower vitamin E levels were less likely to be frail (as assessed by an index of weight loss, low energy, slow gait, low grip strength, and low physical activity; Ble et al., 2006), had higher conduction velocity in peripheral nerves (a slowing of conduction velocity is thought to contribute to decline in muscle strength; Di Iorio et al., 2006), and were less likely to have peripheral arterial disease (Antonelli-Incalzi et al., 2006).

The concept of frailty has become the focus of a growing body of research, stimulated in good measure by Linda Fried, a leader in developing and testing of an operational definition of the concept (Fried et al., 2001). Using data from the WHAS I and II studies, Fried and colleagues have provided confirmation of the internal validity of the component measures, which include poor extremity strength (low grip strength), slow gait, low levels of physical activity, exhaustion and weight loss, and the independent risks for functional disability, institutionalization, and mortality associated with such frailty (Bandeen-Roche et al., 2006). WHAS I and II data have also contributed to understanding of the various physiological processes that appear to contribute to frailty, including documenting the contribution of anemia (Chaves et al., 2005) and, perhaps most surprisingly, of obesity, which was found to be associated with prefrailty and frailty despite the fact that a defining characteristic of frailty is weight loss (Blaum, Xue, Michelon, Semba, and Fried, 2005b). This obesity-frailty association remained significant even with adjustments for multiple conditions associated with frailty (e.g., inflammation burden). Data from the CHS provide parallel evidence linking increased burdens of inflammation to risks for frailty (Walston et al., 2002). Research is needed to determine whether frailty represents an important pathway that underlies known links between a number of the biomarker correlates of frailty (e.g., inflammation, anemia, vitamin E) and disability, morbidity, and mortality outcomes (see review above).


Using available biomarker data, investigators associated with the MacArthur study have been among the leaders in efforts to develop operational indices of allostatic load (AL)—that is, a multisystems measure of physiological dysregulation. Beginning with initial work using a simple count of the number of available biomarkers for which an individual had a value placing them in the top risk quartile, Seeman, McEwen, Singer, Albert, and Rowe (1997b) examined health risks associated with differences in such cumulative AL and demonstrated that higher levels of baseline AL were associated with significantly increased risks for cardiovascular disease, cognitive and physical decline, as well as mortality (Seeman, Singer, Horwitz, and McEwen, 1997b; Seeman Singer, Rowe, and McEwen, 2001). Subsequent work by Karlamangla, Singer, McEwen, Rowe, and Seeman (2002) using canonical correlation techniques demonstrated improved prediction of cognitive and physical decline when the full range of scores was used for each biomarker, and unequal weighting of the different biomarkers was incorporated into the scoring of overall AL. Recent work has also documented that measured change in AL predicts subsequent mortality risk (Karlamangla, Singer, and Seeman, 2006). Stimulated by Singer's work on recursive partitioning (Zhang and Singer, 1999; Singer, Ryff, and Seeman, 2004), recent analyses by Gruenewald, Seeman, Ryff, Karlamangla, and Singer (no date) have demonstrated that particular combinations of these biomarkers (e.g., inflammation and neuroendocrine biomarkers) predict higher versus lower mortality risks over a 12-year period.

A common thread seen in all of the work on AL has been the confirmation of contributions to cumulative health risks from multiple biological systems and the value of taking account of this range of contributions in understanding population variations in burdens of morbidity, disability, and mortality. As illustrated in Seeman et al. (2004a), analyses of AL as a mediator of education effects on mortality risk, the more comprehensive AL index accounted for the largest percentage reduction in the education effect on mortality, with subsets of biomarkers representing cardiovascular, inflammatory, and sympathetic nervous system/hypothalamic-pituitary-adrenal activity, each contributing to this overall effect (Seeman et al., 2004a).

As noted earlier, data from the Wisconsin Longitudinal Study (a cohort of men and women, ages 58-62, approximately a decade younger than the MacArthur study cohort) have provided evidence for relationships between SES and levels of positive social engagement and a similar “count” index of allostatic load (Singer and Ryff, 1999; Seeman et al., 2002). Using the MacArthur Study of Successful Aging protocols, the Social Environment and Biomarkers of Aging Study (SEBAS) (see Chapter 3) collected parallel biological data and provides another comparison to the MacArthur study. Analyses of SEBAS data have yielded intriguing similarities and differences with respect to relationships and allostatic load (weaker in Taiwan; Hu, Wagle, Goldman, Weinstein, and Seeman, 2006), and between social integration and allostatic load (nonsignificant in Taiwan; Seeman, Glei, Goldman, Weinstein, Singer, and Lin, 2004b). The possible importance of sociocultural differences between Asia and the United States represents one area of potentially fruitful future research to better understand how aspects of the social environment impact health.


Two other recent areas of interest in WHAS I and II include examination of possible interactions among biomarkers, with initial work showing cross-sectional relationships between higher serum levels of antioxidants and lower levels of IL-6 (as a marker of inflammation) and longitudinal relationships between initially low antioxidant levels and subsequent increases in IL-6 levels (Walston et al., 2006). A second area of investigation has been to incorporate consideration of genetic information in tracking the factors contributing to observed profiles of biological activity. Initial analyses focused on IL-6 alleles and their relationship to serum IL-6 levels and to decreased muscle strength and frailty. No significant relationships were found for any of these outcomes with any single IL-6 single nucleotide polymorphism (SNP) or any IL-6 haplotype (Walston et al., 2005).

Genetic pleiotropy (in which a gene or set of genes influences multiple traits) could explain the association between biomarkers. A series of studies from the Swedish twin projects have investigated the variance architecture explaining the clustering between biomarkers for cardiovascular disease. Genetic and environmental correlations among the following five serum lipid measures—total cholesterol, HDL cholesterol, triglycerides, and apolipoproteins A-I and B—were analyzed in two different age groups from SATSA. Substantial genetic correlations were found in each age group, although there is no evidence for a single genetic factor common to all five lipids. There were significant age differences in the heritabilities for the various serum lipid levels, and genetic factors seemed to be more important for explaining the covariation between the lipid levels in the younger compared with the older group (Heller, Pedersen, de Faire U, and McClearn, 1994). Further research focused on the sources of clustering among five principal components (BMI, insulin resistance, triglycerides, HDL cholesterol, and systolic blood pressure) of the insulin resistance syndrome (IRS). Results suggest a single set of genetic factors is common to all five components; of particular note was the strong genetic association between BMI and insulin resistance. In contrast, the relationship between only three of the IRS components—triglycerides, insulin resistance, and HDL cholesterol—could be explained by shared sources of individual environmental influences. These findings demonstrating a strong genetic correlation between BMI and insulin resistance raises the question whether behavioral factors that affect both of these phenotypes, such as overeating, act through genetic pathways, perhaps related to control and sensations of satiation, rather than through such environmental factors as availability of food (Hong, Pedersen, Brismar, and deFaire, 1997).


Recent growth in the number of studies incorporating biomarkers into larger population-based surveys has yielded a rapidly growing body of evidence linking various aspects of biological functioning not only to major health outcomes, including cognitive and physical functioning and longevity, but also importantly to individual differences in socioeconomic and other social, psychological, and behavioral characteristics. Findings linking aspects of life situations to major biological risk factors provides important evidence on two fronts. First, it provides validation for various biopsychosocial models of aging and helps elucidate biological pathways through which social, psychological, and behavioral factors affect trajectories of aging and risks for various health outcomes. Second, such evidence provides further support for the potential value of interventions targeting such social, psychological, and behavioral factors as a means of altering underlying biological risk profiles.

A number of ongoing studies will soon provide even richer databases for use in elucidating the complex pathways through which individuals' life experiences and situations affect their health and aging and the biological pathways through which these effects are mediated. For example, current data collection for the MIDUS (Midlife Development in the United States) study will result in a rich set of data on life experiences, including longitudinal data covering the past decade, along with detailed biomarker data and daily diary data for subsets of some 1,500 MIDUS participants. These data will offer significantly improved opportunities to examine a wide variety of hypotheses regarding the role of life experiences (both current and past) in shaping patterns of biological risk and health trajectories. As also outlined in other chapters, data collection for other studies, such as the Health and Retirement Study and the reassessments currently under way for the Taiwan SEBAS study, will offer yet additional national and cross-national data, including socioeconomic, psychosocial, and biological data, that can be used to replicate and extend findings outlined here.

The data and research potential generated from these studies will be greatly enhanced by coordinating efforts such as those undertaken by the Chicago Core on Biomarkers in Population-Based Aging Research (CCBAR) at the University of Chicago-NORC Center on Aging. Through a number of activities, CCBAR provides a central resource to help foster collaboration, exchange information, and promote interdisciplinary research related to biomarker collections in population-based health research and aging, including an interactive website, http://biomarkers.health-studies.org/studydemo.php.

It is instructive to consider the factors that affect the selection of biomarkers for inclusion in these studies. These factors include at least three common and critical considerations—two scientific and one logistical or financial. First, from a scientific standpoint, biomarkers were selected to include those needed to address primary substantive, scientific questions of interest to the study investigators. Second, also from a scientific standpoint, in cases in which there is a presumption from the beginning that the wider community of health researchers will ultimately use these databases to address other, as yet unspecified, questions, there is clearly the additional question of whether additional biomarkers (i.e., other than those already selected based on the substantive interests of study investigators) should be included. Here, considerations generally relate to whether there are major biological systems or processes that are likely to affect health for which data would otherwise be missing.

Perhaps most challenging are a third set of nonscientific considerations that relate to issues of feasibility both with respect to requirements for implementation of protocols to collect and process needed biospecimens and with financial considerations. Perhaps foremost are the logistical constraints imposed by time and handling requirements for obtaining many biological measurements. Examples include (1) the need for phlebotomy to collect venous blood (e.g., when dried blood spots cannot be used), (2) the need for fasted blood samples, which can constrain collection to morning hours, (3) the need for sample collection at specific times due to diurnal rhythms of parameters such as cortisol, necessitating collection at the same time of day for everyone and collection of multiple samples over time, (4) the need for blood or urine samples to be processed within a limited time frame (usually within a couple of hours). The selection of biomarkers for inclusion in a given study will thus necessarily be heavily influenced by what is possible in the context of specific study designs and logistical parameters. For example, the national scope of the HRS study precludes collection of venous blood so dried blood spots are being collected. Selection of biomarkers is thus restricted to those for which assays are available that can use dried blood spots rather than venous blood. By contrast, the MIDUS study is collecting a wide array of biomarkers because participants are being brought to regional clinical research centers (each in a hospital setting) where venous blood can be drawn first thing in the morning (allowing for fasted samples) and where these samples can be processed immediately.

Thus, the final set of biomarkers included in any studies will reflect both the underlying scientific questions investigators seek to address and what is possible given their logistical and financial constraints. Despite the demands and challenges inherent in incorporating biomarkers into social science surveys, continued research development and efforts in the directions presented herein are critical to understanding better the factors that affect patterns of biological aging and trajectories of health at older ages.


  1. Antonelli-Incalzi R, Pedone C, McDermott MM, Bandinelli S, Miniati B, Lova RM, Lauretani F, Ferrucci L. Association between nutrient intake and peripheral artery disease: Results from the InCHIANTI study. Atherosclerosis. 2006;186(1):200–206. [PMC free article: PMC2645647] [PubMed: 16112120]
  2. Bandeen-Roche K, Xue QL, Ferrucci L, Walston J, Guralnik JM, Chaves P, Zeger SL, Fried LP. Phenotype of frailty: Characterization in the Women's Health and Aging Studies. Journals of Gerontology Series A: Biological Sciences and Medical Science. 2006;61(3):262–266. [PubMed: 16567375]
  3. Bassuk SS, Glass TA, Berkman LF. Social disengagement and incident cognitive decline in community-dwelling elderly persons. Annals of Internal Medicine. 1999;131:165–173. [PubMed: 10428732]
  4. Bell B, Rose CL, Damon A. The normative aging study: An interdisciplinary and longitudinal study of health and aging. Aging and Human Development. 1972;3(1):5–17.
  5. Bennett DA, Schneider JA, Buchman AS, et al. The rush memory and aging project: Study design and baseline characteristics of the study cohort. Neuroepidemiology. 2005;25:163–175. [PubMed: 16103727]
  6. Berkman LF, Kawachi I, editors. Social epidemiology. New York: Oxford Press; 2000.
  7. Berkman LF, Seeman TE, Albert MA, et al. High, usual and impaired functioning in community-dwelling older men and women: Findings from the MacArthur Foundation Research Network on Successful Aging. Journal of Clinical Epidemiology. 1993;46:1129–1140. [PubMed: 8410098]
  8. Blaum CS, Volpato S, Cappola AR, Chaves P, Xue QL, Guralnik JM, Fried LP. Diabetes, hyperglycaemia, and mortality in disabled older women: The Women's Health and Ageing Study I. Diabetic Medicine. 2005;22(5):543–550. [PubMed: 15842507]
  9. Blaum CS, Xue QL, Michelon E, Semba RD, Fried LP. The association between obesity and the frailty syndrome in older women: The Women's Health and Aging Studies. Journal of the American Geriatrics Society. 2005;53(6):1069–1070. [PubMed: 15935013]
  10. Ble A, Cherubini A, Volpato S, Bartali B, Walston JD, Windham BG, Bandinelli S, Lauretani F, Guralnik JM, Ferrucci L. Lower plasma vitamin E levels are associated with the frailty syndrome: The InCHIANTI study. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2006;61(3):278–283. [PubMed: 16567378]
  11. Boomsma DI, Vink JM, van Beijsterveldt TC, de Geus EJ, Beem AL, Mulder EJ, Derks EM, Riese H, Willemsen GA, Bartels M, van den Berg M, Kupper NH, Polderman TJ, Posthuma D, Rietveld MJ, Stubbe JH, Knol LI, Stroet T, van Baal GC. Netherlands Twin Register: A focus on longitudinal research. Twin Research and Human Genetics. 2002;5(5):401–406. [PubMed: 12537867]
  12. Boomsma DI, Willemsen G, Dolan CV, Hawkley LC, Cacioppo JT. Genetic and environmental contributions to loneliness in adults: The Netherlands twin register study. Behavioral Genetics. 2005;35(6):745–752. [PubMed: 16273322]
  13. Boomsma DI, Cacioppo JT, Slagboom PE, Posthuma D. Genetic linkage and association analysis for loneliness in Dutch twin and sibling pairs points to a region on chromosome 12q23-24. Behavioral Genetics. 2006;36(1):137–146. Epub 2005 Dec 24. [PubMed: 16378171]
  14. Bretsky P, Guralnik J, Launer L, Albert MA, Seeman TE. The role of APOE-ɛ4 in longitudinal cognitive decline in an elderly cohort: MacArthur studies of successful aging. Neurology. 2003;60:1077–1081. [PubMed: 12682309]
  15. Burg MM, Seeman TE. Families and health: The negative side of social ties. Annals of Behavioral Medicine. 1994;16:109–115.
  16. Butler J, Rodondi N, Zhu Y, Figaro K, Fazio S, Vaughan DE, Satterfield S, Newman AB, Goodpaster B, Bauer DC, Holvoet P, Harris TB, de Rekeneire N, Rubin S, Ding J, Kritchevsky SB., Health ABC Study. Metabolic syndrome and the risk of cardiovascular disease in older adults. Journal of the American College of Cardiology. 2006;47(8):1595–1602. [PubMed: 16630996]
  17. Butler RN, Sprott R, Warner H, Bland J, Feuers R, Forster M, Fillit H, Harman M, Hewitt M, Hyman M, Johnson K, Kligman E, McClearn G, Nelson J, Richardson A, Sonntag W, Weindruch R, Wolf N. Biomarkers of aging: From primitive organisms to humans. Journal of Gerontology: Biological Sciences. 2004;59A:560–567. [PubMed: 15215265]
  18. Cacioppo JT, Hughes ME, Waite LJ, Hawkley LC, Thisted RA. Loneliness as a specific risk factor for depressive symptoms: Cross-sectional and longitudinal analyses. Psychology and Aging. 2006;21(1):140–151. [PubMed: 16594799]
  19. Cappola AR, Bandeen-Roche K, Wand GS, Volpato S, Fried LP. Association of IGF-I levels with muscle strength and mobility in older women. Journal of Clinical Endocrinology & Metabolism. 2001;86(9):4139–4146. [PubMed: 11549640]
  20. Caspi A, Sugden K, Moffitt TE, et al. Influence of life stress on depression: Moderation by a polymorphism in the 5-HTT gene. Science. 2003;301:386–389. [PubMed: 12869766]
  21. Cesari M, Penninx BW, Newman AB, Kritchevsky SB, Nicklas BJ, Sutton-Tyrrell K, Tracy RP, Rubin SM, Harris TB, Pahor M. Inflammatory markers and cardiovascular disease (The Health, Aging, and Body Composition [Health ABC] Study) The American Journal of Cardiology. 2003;92(5):522–528. [PubMed: 12943870]
  22. Cesari M, Penninx BW, Newman AB, Kritchevsky SB, Nicklas BJ, Sutton-Tyrrell K, Rubin SM, Ding J, Simonsick EM, Harris TB, Pahor M. Inflammatory markers and onset of cardiovascular events: Results from the Health ABC study. Circulation. 2003;108(19):2317–2322. [PubMed: 14568895]
  23. Cesari M, Penninx BW, Pahor M, Lauretani F, Corsi AM, Rhys Williams G, Guralnik JM, Ferrucci L. Inflammatory markers and physical performance in older persons: The InCHIANTI study. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2004;59(3):242–248. [PubMed: 15031308]
  24. Cesari M, Penninx BW, Lauretani F, Russo CR, Carter C, Bandinelli S, Atkinson H, Onder G, Pahor M, Ferrucci L. Hemoglobin levels and skeletal muscle: Results from the InCHIANTI study. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2004;59(3):249–254. [PubMed: 15031309]
  25. Cesari M, Pahor M, Lauretani F, Penninx BW, Bartali B, Russo R, Cherubini A, Woodman R, Bandinelli S, Guralnik JM, Ferrucci L. Bone density and hemoglobin levels in older persons: Results from the InCHIANTI study. Osteoporosis International. 2005;16(6):691–699. [PubMed: 15455197]
  26. Chaves PH, Semba RD, Leng SX, Woodman RC, Ferrucci L, Guralnik JM, Fried LP. Impact of anemia and cardiovascular disease on frailty status of community-dwelling older women: The Women's Health and Aging Studies I and II. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2005;60(6):729–735. [PubMed: 15983175]
  27. Cherubini A, Martin A, Andres-Lacueva C, Di Iorio A, Lamponi M, Mecocci P, Bartali B, Corsi A, Senin U, Ferrucci L. Vitamin E levels, cognitive impairment, and dementia on older persons: The InCHIANTI study. Neurobiology of Aging. 2005;26(7):987–994. [PubMed: 15748776]
  28. Crimmins EM, Alley D, Reynolds SL, Johnston M, Karlamangla A, Seeman T. Changes in biological markers of health: Older Americans in the 1990s. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2005;60(11):1409–1413. [PubMed: 16339326]
  29. de Quervain DJF, Henke K, Aerni A, Coluccia D, Wollmer MA, Hock C, et al. A functional genetic variation of the 5-HT2A receptor affects human memory. Nature Neuroscience. 2003;6(11):1141–1142. [PubMed: 14566344]
  30. Di Iorio A, Cherubini A, Volpato S, Sparvieri E, Lauretani F, Franceschi C, Senin U, Abate G, Paganelli R, Martin A, Andres-Lacueva C, Ferrucci L. Markers of inflammation, vitamin E and peripheral nervous system function: The InCHIANTI study. Neurobiology of Aging. 2006;27(9):1280–1288. [PMC free article: PMC2646090] [PubMed: 16112778]
  31. Ding J, Nieto FJ, Beauchamp NJ, Longstreth WT Jr, Manolio TA, Hetmanski JB, Fried LP. A prospective analysis of risk factors for white matter disease in the brain stem: The Cardiovascular Health Study. Neuroepidemiology. 2003;22(5):275–282. [PubMed: 12902622]
  32. Elosua R, Bartali B, Ordovas JM, Corsi AM, Lauretani F, Ferrucci L., InCHIANTI Investigators. Association between physical activity, physical performance, and inflammatory biomarkers in an elderly population: The InCHIANTI study. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2005;60(6):760–767. [PubMed: 15983180]
  33. Eley TC, Sugden K, Corsico A, Gregary AM, Sham P, McGuffin P, Plomin R, Craig IW. Gene-environment interaction analysis of serotonin system markers with adolescent depression. Molecular Psychiatry. 2004;9:908–915. [PubMed: 15241435]
  34. Epel ES, Blackburn EH, Lin J, Dhabhar FS, Adler NE, Morrow JD, Cawthon RM. Accelerated telomere shortening in response to life stress. Proceedings of the National Academy of Sciences USA. 2004;101(49):17312–17315. Epub 2004 Dec 1. [PMC free article: PMC534658] [PubMed: 15574496]
  35. Ewbank DC. Mortality differences by APOE genotype estimated from demographic synthesis. Genetic Epidemiology. 2002;(22):146–155. [PubMed: 11788960]
  36. Ferrucci L, Penninx BW, Volpato S, Harris TB, Bandeen-Roche K, Balfour J, Leveille SG, Fried LP, Guralnik JM. Change in muscle strength explains accelerated decline of physical function in older women with high interleukin-6 serum levels. Journal of the American Geriatrics Society. 2002;50(12):1947–1954. [PubMed: 12473005]
  37. Fratiglioni L, Wang HX, Ericsson K, Maytan M, Winblad B. Influence of social network on occurrence of dementia: A community-based longitudinal study. Lancet. 2000;355:1315–1319. [PubMed: 10776744]
  38. Fried LP, Kronmal RA, Newman AB, Bild DE, Mittelmark MB, Polak JF, Robbins JA, Gardin JM. Risk factors for 5-year mortality in older adults: The Cardiovascular Health Study. Journal of the American Medical Association. 1998;279(8):585–592. [PubMed: 9486752]
  39. Fried LP, Bandeen-Roche K, Chaves PHM, Johnson BA. Preclinical mobility disability predicts incident mobility disability in older women. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2000;55A(1):M43–M52. [PubMed: 10719772]
  40. Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, Seeman T, Tracy R, Kop WJ, Burke G, McBurnie MA., Cardiovascular Health Study Collaborative Research Group. Frailty in older adults: Evidence for a phenotype. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2001;56(3):M146–M156. [PubMed: 11253156]
  41. Friedman EM, Hayney MS, Love GD, Urry HL, Rosenkranz MA, Davidson RJ, Singer BH, Ryff CD. Social relationships, sleep quality, and interleukin-6 in aging women. Proceedings of the National Academy of Sciences, USA. 2005;102(51):18757–18762. [PMC free article: PMC1317967] [PubMed: 16339311]
  42. Gatz M, Pedersen NL, Berg S, Johansson B, et al. Heritability for Alzheimer's disease: The study of dementia in Swedish twins. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 1997;(52A):M117–M125. [PubMed: 9060980]
  43. Gillespie NA, Whitfield JB, Williams B, Heath AC, Martin N. The relationship between stressful life events, the serotonin transporter (5-HTTLPR) genotype, and major depression. Psychological Medicine. 2005;35:101–111. [PubMed: 15842033]
  44. Gottesman II, Gould TD. The endophenotype concept in psychiatry: Etymology and strategic intentions. American Journal of Psychiatry. 2003;160(4):636–645. [PubMed: 12668349]
  45. Grabe HJ, Lange M, Wolff B, Volzke H, Lucht M, Freyberger HJ, John U, Cascorbi I. Mental and physical distress is modulated by a polymorphism in the 5-HT transporter gene interacting with social stressors and chronic disease burden. Molecular Psychiatry. 2005;10:220–224. [PubMed: 15263905]
  46. Gruenewald TL, Seeman TE, Ryff CD, Karlamangla AS, Singer BH. Early warning biomarkers: Combinations predictive of later life mortality. Proceedings of the National Academy of Sciences, USA. 2006;103(3i8):14158–14163. [PMC free article: PMC1599928] [PubMed: 16983099]
  47. Harris JR. Research on environmental effects in genetic studies of aging: Introduction. The Journal of Gerontology B: Psychological and Social Sciences. 2005;60(Spec No. 1):5–6. [PubMed: 15863704]
  48. Harris JR. Genetics, social behaviors, social environments, and aging. Twin Research and Human Genetics. 2007;10(2):235–240. [PubMed: 17564513]
  49. Hawkley LC, Cacioppo JT. Loneliness and pathways to disease. Brain, Behavior, and Immunity. 2003;17:S98–S105. [PubMed: 12615193]
  50. Hawkley LC, Masi CM, Berry JD, Cacioppo JT. Loneliness is a unique predictor of age-related differences in systolic blood pressure. Psychology and Aging. 2006;21:152–164. [PubMed: 16594800]
  51. Heller DA, de Faire U, Pedersen NL, Dahlen G, McClearn GE. Genetic and environmental influences on serum lipid levels in twins. New England Journal of Medicine. 1993;328(16):1150–1156. [PubMed: 8455681]
  52. Heller DA, Pedersen NL, de Faire U, McClearn GE. Genetic and environmental correlations among serum lipids and apolipoproteins in elderly twins reared together and apart. American Journal of Human Genetics. 1994;(6):1255–1267. [PMC free article: PMC1918425] [PubMed: 7977387]
  53. Hong Y, Dahlen G, Pedersen N, Heller D, McClearn GE, deFaire U. Potential environmental effects on adult lipoprotein(a) levels: Results from Swedish twins. Atherosclerosis. 1995;17:295–304. [PubMed: 8801875]
  54. Hong Y, Pedersen NL, Brismar K, de Faire U. Genetic and environmental architecture of the features of the insulin-resistance syndrome. American Journal of Human Genetics. 1997;60(1):143–152. [PMC free article: PMC1712553] [PubMed: 8981957]
  55. Hong Y, Pedersen NL, Egberg N, de Faire U. Genetic effects for plasma factor VII levels independent of and in common with triglycerides. Journal of Thrombosis and Haemostasis. 1999;81(3):382–386. [PubMed: 10102465]
  56. House JS, Landis KR, Umberson D. Social relationships and health. Science. 1988;241:540–545. [PubMed: 3399889]
  57. House JS, Lepkowski JM, Kinney AM, Mero RP, Kessler RC, Herzog AR. The social stratification of aging and health. Journal of Health and Social Behavior. 1994;35:213–234. [PubMed: 7983335]
  58. Hu P, Reuben DB, Crimmins E, Harris TB, Huang MH, Seeman TE. The effects of serum beta-carotene level and burden of inflammation on all-cause mortality in high-functioning older persons: MacArthur Studies of Successful Aging. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2005;59:849–854. [PubMed: 15345737]
  59. Hu P, Wagle N, Goldman N, Weinstein M, Seeman TE. The associations between socioeconomic status, allostatic load and measures of health in older Taiwanese persons: Taiwan social environment and biomarkers of aging study. Journal of Biosocial Science. 2006;20:1–12. [PubMed: 17052381]
  60. Huang TL, Zandi PP, Tucker KL, Fitzpatrick AL, Kuller LH, Fried LP, Burke GL, Carlson MC. Benefits of fatty fish on dementia risk are stronger for those without APOE epsilon4. Neurology. 2005;65(9):1409–1414. [PubMed: 16275829]
  61. Iliadou A, Lichtenstein P, de Faire U, Pedersen NL. Variation in genetic and environmental influences in serum lipid and apolipoprotein levels across the lifespan in Swedish male and female twins. American Journal of Medical Genetics. 2001;102(1):48–58. [PubMed: 11471172]
  62. Iliadou A, Sneider H. Genetic epidemiological approaches in the study of risk factors for cardiovascular disease. European Journal of Epidemiology. 2004;19:209–217. [PubMed: 15117113]
  63. Institute of Medicine. Genes, behavior, and the social environment. Washington, DC: The National Academies Press; 2006.
  64. Jansson M, Gatz M, Berg S, Johansson B, Malmberg B, McClearn GE, Schalling M, Pedersen NL. Association between depressed mood in the elderly and a 5-HTR2A gene variant. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics. 2003;120(1):79–84. [PubMed: 12815744]
  65. Jenny NS, Arnold AM, Kuller LH, Sharrett AR, Fried LP, Psaty BM, Tracy RP. Soluble intracellular adhesion molecule-1 is associated with cardiovascular disease risk and mortality in older adults. Journal of Thrombosis and Haemostasis. 2006;4(1):107–113. [PubMed: 16409459]
  66. Kado DM, Karlamangla AS, Huang MH, Troen A, Rowe JW, Selhub J, Seeman TE. Homocysteine versus the vitamins folate, B6 and B12 as predictors of cognitive function and decline in older high functioning adults: MacArthur Studies of Successful Aging. American Journal of Medicine. 2005;118:161–167. [PubMed: 15694902]
  67. Karlamangla AS, Singer BH, McEwen BS, Rowe JW, Seeman TE. Allostatic lead as a predictor of functional decline: MacArthur Studies of Successful Aging. Journal of Clinical Epidemiology. 2002;55(7):696–710. [PubMed: 12160918]
  68. Karlamangla AS, Singer BS, Reuben D, Seeman TE. Increase in serum non-high-density lipoprotein cholesterol may be beneficial in some high-functioning older adults: MacArthur Studies of Successful Aging. Journal of the American Geriatrics Society. 2004;52:487–494. [PubMed: 15066061]
  69. Karlamangla AS, Singer B, Greendale G, Seeman T. Increase in epinephrine excretion is associated with cognitive decline in elderly men: MacArthur Studies of Successful Aging. Psychoneuroendocrinology. 2005;30(5):453–460. [PubMed: 15721057]
  70. Karlamangla AS, Singer BH, Chodosh J, McEwen BS, Seeman TE. Urinary cortisol excretion as a predictor of incident cognitive impairment. Neurobiology of Aging. 2005;26S:S80–S84. [PubMed: 16280187]
  71. Karlamangla AS, Singer BH, Seeman TE. Reduction in allostatic load in older adults is associated with lower all-cause mortality risk: MacArthur Studies of Successful Aging. Psychosomatic Medicine. 2006;68:500–507. [PubMed: 16738085]
  72. Kaufman J, Yang BZ, Douglas-Palumberi H, Houshyar S, Lipshitz D, Krystal J, Gelernter J. Social supports and serotonin transporter gene moderate depression in maltreated children. Proceedings of the National Academy of Sciences, USA. 2004;101:17316–17321. [PMC free article: PMC534414] [PubMed: 15563601]
  73. Kendler KS, Kuhn JW, Vittum J, Prescott CA, Riley B. The interaction of stressful life events and a serotonin transporter polymorphism in the prediction of episodes of major depression: A replication. Archives of General Psychiatry. 2005;62:529–535. [PubMed: 15867106]
  74. Kiecolt-Glaser JK, Loving TJ, Stowell JR, Malarkey WB, Lemeshow S, Dickinson SL, Glaser R. Hostile marital interactions, proinflammatory cytokine production, and wound healing. Archives of General Psychiatry. 2005;62:1377–1384. [PubMed: 16330726]
  75. Koster A, Penninx BW, Bosma H, Kempen GI, Harris TB, Newman AB, Rooks RN, Rubin SM, Simonsick EM, van Eijk JT, Kritchevsky SB. Is there a biomedical explanation for socioeconomic differences in incident mobility limitation? Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2005;60(8):1022–1027. [PubMed: 16127107]
  76. Koster A, Penninx BW, Bosma H, Kempen GI, Newman AB, Rubin SM, Satterfield S, Atkinson HH, Ayonayon HN, Rosano C, Yaffe K, Harris TB, Rooks RN, Van Eijk JT, Kritchevsky SB. Socioeconomic differences in cognitive decline and the role of biomedical factors. Annals of Epidemiology. 2005;15(8):564–571. [PubMed: 15922627]
  77. Koster A, Bosma H, Penninx BW, Newman AB, Harris TB, van Eijk JT, Kempen GI, Simonsick EM, Johnson KC, Rooks RN, Ayonayon HN, Rubin SM, Kritchevsky SB., Health ABC Study. Association of inflammatory markers with socioeconomic status. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2006;61(3):284–290. [PubMed: 16567379]
  78. Kubzansky LD, Kawachi I, Sparrow D. Socioeconomic status, hostility, and risk factor clustering in the Normative Aging Study: Any help from the concept of allostatic load? Annals of Behavioral Medicine. 1999;21(4):330–338. [PubMed: 10721441]
  79. Lichtenstein P, de Faire U, Floderus B, Svartenren M, Svedberg P, Pedersen NL. The Swedish Twin Registry: A unique resource for clinical, epidemiological, and genetic studies. Journal of International Medical Research. 2002;(252):184–205. [PubMed: 12270000]
  80. Loucks EB, Berkman LF, Gruenewald TL, Seeman TE. Relation of social integration to inflammatory marker concentrations in men and women 70-79 years. American Journal of Cardiology. 2006;97(7):1010–1016. [PubMed: 16563907]
  81. McClearn GE, Johansson B, Berg S, Pedersen NL, Ahern F, Petrill SA, Plomin R. Substantial genetic influence on cognitive abilities in twins 80 or more years old. Science. 1997;276(5318):1560–1563. [PubMed: 9171059]
  82. McDermott MM, Guralnik JM, Corsi A, Albay M, Macchi C, Bandinelli S, Ferrucci L. Patterns of inflammation associated with peripheral arterial disease: The InCHIANTI study. American Heart Journal. 2005;150(2):276–281. [PubMed: 16086930]
  83. McGuire S, Clifford J. Genetic and environmental contributions to loneliness in children. Psychological Sciences. 2000;11:487–491. [PubMed: 11202494]
  84. Moceri VM, Kukull WA, Emanual I, van Belle G, Starr JR, Schellenberg GD, McCormick WC, Bowen JD, Teri L, Larson EB. Using census data and birth certificates to reconstruct the early-life socioeconomic environment and the relation to the development of Alzheimer's disease. Epidemiology. 2001;(12):383–389. [PubMed: 11416775]
  85. National Research Council; Commission on Behavioral and Social Sciences and Education. Cells and surveys: Should biological measures be included in social science research? Committee on PopulationFinch CE, Vaupel JW, Kinsella K, editors. Washington, DC: National Academy Press; 2001. [PubMed: 23166967]
  86. National Research Council; Commission on Behavioral and Social Sciences and Education. Between Zeus and the salmon: The biodemography of longevity. Committee on PopulationWachter KW, Finch CE, editors. Washington, DC: National Academy Press; 1997. [PubMed: 22973581]
  87. National Research Council; Board on Behavioral, Cognitive, and Sensory Sciences. Commission on Behavioral and Social Sciences and Education. New horizons in health: An integrative approach. Committee on Future Directions for Behavioral and Social Sciences Research at the National Institutes of HealthSinger BH, Ryff CD, editors. Washington, DC: National Academy Press; 2001.
  88. Niaura R, Banks SM, Ward KD, Stoney C.M, Spiro A III, Aldwin CM, Landsberg L, Weiss ST. Hostility and the metabolic syndrome in older males: The normative aging study. Psychosomatic Medicine. 2000;62(1):7–16. [PubMed: 10705906]
  89. Nilsson SE, Evrin PE, Tryding N, Berg S, McClearn G, Johansson B. Biochemical values in persons older than 82 years of age: Report from a population-based study of twins. Scandinavian Journal of Clinical and Laboratory Investigation. 2003;63(1):1–13. [PubMed: 12729064]
  90. Nilsson SE, Takkinen S, Tryding N, Evrin PE, Berg S, McClearn G, Johansson B. Association of biochemical values with morbidity in the elderly: A population-based Swedish study of persons aged 82 or more years. Scandinavian Journal of Clinical and Laboratory Investigation. 2003;63(7-8):457–466. [PubMed: 14743954]
  91. Onder G, Penninx BW, Cesari M, Bandinelli S, Lauretani F, Bartali B, Gori AM, Pahor M, Ferrucci L. Anemia is associated with depression in older adults: Results from the InCHIANTI study. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2005;60(9):1168–1172. [PubMed: 16183958]
  92. Payton M, Riggs KM, Spiro A III, Weiss ST, Hu H. Relations of bone and blood lead to cognitive function: The VA Normative Aging Study. Neurotoxicology and Teratology. 1998;20(1):19–27. [PubMed: 9511166]
  93. Pedersen NL, McClearn GE, Plomin R, Nesselroade JR, Berg S, DeFaire U, Mortimer JA, et al. The Swedish Adoption Twin Study of Aging: An update. Acta Geneticae Medicae et Gemellologiae. 1991;(40):7–20. [PubMed: 1950353]
  94. Penninx BW, Guralnik JM, Ferrucci L, Fried LP, Allen RH, Stabler SP. Vitamin B(12) deficiency and depression in physically disabled older women: Epidemiologic evidence from the Women's Health and Aging Study. American Journal of Psychiatry. 2000;157(5):715–721. [PubMed: 10784463]
  95. Penninx BW, Pahor M, Cesari M, Corsi AM, Woodman RC, Bandinelli S, Guralnik JM, Ferrucci L. Anemia is associated with disability and decreased physical performance and muscle strength in the elderly. Journal of the American Geriatric Society. 2005;52(5):719–724. [PubMed: 15086651]
  96. Psaty BM, Anderson M, Kronmal RA, Tracy RP, Orchard T, Fried LP, Lumley T, Robbins J, Burke G, Newman AB, Furberg CD. The association between lipid levels and the risk of incident myocardial infarction, stroke, and total mortality: The Cardiovascular Health Study. Journal of the American Geriatric Society. 2004;52(10):1639–1647. [PubMed: 15450039]
  97. Rantanen T, Guralnik JM, Sakari-Rantala R, Leveille S, Simonsick EM, Ling S, Fried LP. Disability, physical activity, and muscle strength in older women: The Women's Health and Aging Study. Archives of Physical Medicine and Rehabilitation. 1999;80(2):130–135. [PubMed: 10025485]
  98. Ray AL, Semba RD, Walston J, Ferrucci L, Cappola AR, Ricks MO, Xue QL, Fried LP. Low serum selenium and total carotenoids predict mortality among older women living in the community: The Women's Health and Aging Studies. Journal of Nutrition. 2006;136(1):172–176. [PubMed: 16365078]
  99. Reuben DB, Cheh AI, Harris TB, Ferrucci L, Rowe JW, Tracy RP, Seeman TE. Peripheral blood markers of inflammation predict mortality and functional decline among high functioning community-dwelling older persons. Journal of the American Geriatric Society. 2002;50(4):638–644. [PubMed: 11982663]
  100. Reuben DB, Judd-Hamilton L, Harris TB, Seeman TE. The associations between physical activity and inflammatory markers in high functioning older persons: MacArthur Studies of Successful Aging. Journal of the American Geriatrics Society. 2003 August;51(8):1125–1130. [PubMed: 12890077]
  101. Reynolds CA, Jansson M, Gatz M, Pedersen NL. Longitudinal change in memory performance associated with HTR2A polymorphism. Neurobiology of Aging. 2006;27(1):150–154. [PubMed: 16298250]
  102. Reynolds CA, Prince JA, Feuk L, Brookes AJ, Gatz M, Pedersen NL. Longitudinal memory performance during normal aging: Twin association models of APOE and other Alzheimer candidate genes. Behavioral Genetics. 2006;36(2):185–194. Epub 2006 Jan 10. [PubMed: 16402284]
  103. Reynolds CA, Gatz M, Berg S, Pedersen NL. Genotype-environment interactions: Cognitive aging and social factors. Twin Research and Human Genetics. 2007;10(2):241–254. [PubMed: 17564514]
  104. Rutter M, Moffitt TE, Caspi A. Gene-environment interplay and psychopathology: Multiple varieties but real effects. Journal of Child Psychology and Psychiatry. 2006;47(3-4):226–261. [PubMed: 16492258]
  105. Ryff CD, Singer BH, editors. Emotion, social relationships, and health. New York: Oxford University Press; 2001.
  106. Ryff C, Singer BH. Social environments and the genetics of aging: Advancing knowledge of protective health mechanisms. The Journal of Gerontology B: Psychological and Social Sciences. 2005;60:12–23. [PubMed: 15863706]
  107. Ryff CD, Singer BH, Dienberg Love G. Positive health: Connecting well-being with biology. Philosophical Transactions of the Royal Society B: Biological Sciences. 2004;359:1383–1394. [PMC free article: PMC1693417] [PubMed: 15347530]
  108. Ryff CD, Dienberg Love G, Urry HL, Muller D, Rosenkranz MA, Friedman EM, Davidson RJ, Singer B. Psychological well-being and ill-being: Do they have distinct or mirrored biological correlates? Psychotherapy and Psychosomatics. 2006;75(2):85–95. [PubMed: 16508343]
  109. Saxton J, Lopez OL, Ratcliff G, Dulberg C, Fried LP, Carlson MC, Newman AB, Kuller L. Preclinical Alzheimer disease: Neuropsychological test performance 1.5 to 8 years prior to onset. Neurology. 2004;63(12):2341–2347. [PubMed: 15623697]
  110. Seeman TE. Social ties and health. Annals of Epidemiology. 1996;6:442–451. [PubMed: 8915476]
  111. Seeman TE, Berkman LF, Blazer D, Rowe J. Social ties and support and neuroendocrine function: MacArthur Studies of Successful Aging. Annals of Behavioral Medicine. 1994;16:95–106.
  112. Seeman TE, Berkman LF, Kohout F, LaCroix A, Glynn R, Blazer D. Intercommunity variations in the association between social ties and mortality in the elderly: A comparative analysis of three communities. Annals of Epidemiology. 1993;3:325–335. [PubMed: 8275207]
  113. Seeman TE, Bruce ML, McAvay G. Social network characteristics and onset of ADL disability: MacArthur Studies of Successful Aging. Journal of Gerontology: Psychological Sciences. 1996;51B:S191–S200. [PubMed: 8673648]
  114. Seeman TE, Crimmins E, Bucur A, Huang MH, Singer B, Gruenewald T, Berkman LF, Reuben DB. Cumulative biological risk and socioeconomic differences in mortality: MacArthur Studies of Successful Aging. Social Science and Medicine. 2004;58:1985–1997. [PubMed: 15020014]
  115. Seeman TE, Glei D, Goldman N, Weinstein M, Singer B, Lin YH. Social relationships and allostatic load in Taiwanese elderly and near elderly. Social Science and Medicine. 2004;59:2245–2257. [PubMed: 15450701]
  116. Seeman TE, Huang MH, Bretsky P, Crimmins E, Launer L, Guralnik JM. Education and APOE-ɛ4 in longitudinal cognitive decline: MacArthur Studies of Successful Aging. Journal of Gerontology: Psychological Sciences. 2005;60(2):P74–P83. [PubMed: 15746021]
  117. Seeman TE, Kaplan GA, Knudsen L, Cohen R, Guralnik J. Social ties and mortality in the elderly: A comparative analysis of age-dependent patterns of association. American Journal of Epidemiology. 1987;126:714–723. [PubMed: 3631060]
  118. Seeman TE, Lusignolo T, Berkman L, Albert M. Social environment characteristics and patterns of cognitive aging: MacArthur Studies of Successful Aging. Health Psychology. 2001;20:243–255. [PubMed: 11515736]
  119. Seeman TE, McEwen BS. Impact of social environment characteristics on neuroendocrine regulation. Psychosomatic Medicine. 1996;58(5):459–471. [PubMed: 8902897]
  120. Seeman TE, McEwen BS, Singer B, Albert M, Rowe JW. Increase in urinary cortisol excretion and declines in memory: MacArthur Studies of Successful Aging. Journal of Clinical Endocrinology & Metabolism. 1997;82:2458–2465. [PubMed: 9253318]
  121. Seeman TE, Merkin S, Crimmins E, Koretz B, Karlamangla A. Socioeconomic status, ethnicity, and cumulative biological risk profiles in a national sample of U.S. adults: NHANES III. no date. manuscript in preparation.
  122. Seeman TE, Singer B, Horwitz R, McEwen BS. The price of adaptation— allostatic load and its health consequences: MacArthur Studies of Successful Aging. Archives of Internal Medicine. 1997;157:2259–2268. [PubMed: 9343003]
  123. Seeman TE, Singer B, Rowe J, McEwen B. Exploring a new concept of cumulative biological risk—allostatic load and its health consequences: MacArthur Studies of Successful Aging. Proceedings of the National Academy of Sciences, USA. 2001;98(8):4770–4775. [PMC free article: PMC31909] [PubMed: 11287659]
  124. Seeman TE, Singer B, Ryff C, Levy-Storms L. Psychosocial factors and the development of allostatic load. Psychosomatic Medicine. 2002 May/June;64:395–406. [PubMed: 12021414]
  125. Shadlen MF, Larson EB, Wang L, Phelan EA, McCormick WC, Jolley L, Teri L, van Belle G. Education modifies the effect of apolipoprotein epsilon 4 on cognitive decline. Neurobiology of Aging. 2005;26(1):17–24. [PubMed: 15585342]
  126. Simonsick EM, Maffeo CE, Rogers SK, Skinner EA, Davis D, Guralnik JM, Fried LP. Methodology and feasibility of a home-based examination in disabled older women: The Women's Health and Aging Study. Journals of Gerontology. 1997;52A(5):M264–M274. [PubMed: 9310080]
  127. Singer B, Ryff C. Hierarchies of life histories and associated health risks. Annals of New York Academy of Sciences. 1999;896:96–115. [PubMed: 10681891]
  128. Singer B, Ryff C, Seeman TE. Allostasis, homeostasis, and the cost of physiological adaptation. In: Schulkin J, editor. Operationalizing allostatic load. Cambridge, England: Cambridge University Press; 2004. pp. 113–149.
  129. Sneider H, van Doornen LJ, Boomsma DI. The age dependency of gene expression for plasma lipids, lipoproteins, and apolipoproteins. American Journal of Human Genetics. 1997;(60):638–650. [PMC free article: PMC1712506] [PubMed: 9042925]
  130. Snowdon DA, Kemper SJ, Mortimer J.A, Wekstein DR, Markesbery WR. Linguistic ability in early life and cognitive function and Alzheimer's disease in late life: Findings from the Nun study. Journal of the American Medical Association. 1996;257(7):528–532. [PubMed: 8606473]
  131. Surtees PG, Wainwright NW, Willis-Owen SA, Luben R, Day NE, Flint J. Social adversity, the serotonin transporter (5-HTTLPR) polymorphism, and major depressive disorder. Biological Psychology. 2006;59:224–229. [PubMed: 16154545]
  132. Tang MX, Stern Y, Marder K, Bell K, Gurland B, Lantigua R, Andrews H, Feng L, Tycko B, Mayeux R. The APOE-epsilon4 allele and the risk of Alzheimer's disease among African Americans, Whites, and Hispanics. Journal of the American Medical Association. 1998;11:751–575. [PubMed: 9508150]
  133. Taylor SE, Repetti RL, Seeman T. Health psychology: What is an unhealthy environment and how does it get under the skin? Annual Review of Psychology. 1997;48:411–447. [PubMed: 9046565]
  134. Todaro JF, Con A, Niaura R, Spiro A III, Ward KD, Roytberg A. Combined effect of the metabolic syndrome and hostility on the incidence of myocardial infarction (the Normative Aging Study) American Journal of Cardiology. 2005;96(2):221–226. [PubMed: 16018847]
  135. Tomaka J, Thompson S, Palacios RJ. High homocysteine and low B vitamins predict cognitive decline in aging men: The Veterans Affairs Normative Aging Study. American Journal of Clinical Nutrition. 2005;82(3):627–635. [PubMed: 16155277]
  136. Tomaka J, Thompson S, Palacios RJ. The relation of social isolation, loneliness, and social support to disease outcomes among the elderly. Aging Health. 2006 June)18(3):359–384. [PubMed: 16648391]
  137. Tucker KL, Qiao N, Scott T, Rosenbert I, Spiro A III. High homocysteine and low B vitamins predict cognitive decline in aging men: The Veterans Affairs normative aging study. American Journal of Clinical Nutrition. 2005;82(3):627–635. [PubMed: 16155277]
  138. Uchino BN, Cacioppo JT, Kiecolt-Glase JK. The relationship between social support and physiological processes: A review with emphasis on underlying mechanisms and implications for health. Psychological Bulletin. 1996;119:488–531. [PubMed: 8668748]
  139. Visser M, Pahor M, Taaffe DR, Goodpaster BH, Simonsick EM, Newman AB, Nevitt M, Harris TB. Relationship of interlukin-6 and tumor necrosis factor-alpha with muscle mass and muscle strength in elderly men and women: The Health ABC Study. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2002;57(5):M326–M332. [PubMed: 11983728]
  140. Volpato S, Guralnik JM, Fried LP, Remaley AT, Cappola AR, Launer LJ. Serum thyroxine level and cognitive decline in euthyroid older women. Neurology. 2002;58(7):1055–1061. [PubMed: 11940692]
  141. von Kanel R, Dimsdale JE, Ancoli-Israel S, Mills PJ, Patterson TL, McKibbin CL, Archuleta C, Grant I. Poor sleep is associated with higher plasma proinflammatory cytokine interleukin-6 and procoagulant marker fibrin D-dimer in older caregivers of people with Alzheimer's disease. Journal of the American Geriatrics Society. 2006;54(3):431–437. [PubMed: 16551309]
  142. Walston J, McBurnie MA, Newman A, Tracy RP, Kop WJ, Hirsch CH, Gottdiener J, Fried LP. Frailty and activation of the inflammation and coagulation systems with and without clinical comorbidities: Results from the Cardiovascular Health Study. Archives of Internal Medicine. 2002;162(20):2333–2341. [PubMed: 12418947]
  143. Walston J, Arking DE, Fallin D, Li T, Beamer B, Xue Q, Ferrucci L, Fried LP, Chakravarti A. IL-6 gene variation is not associated with increased serum levels of IL-6, muscle, weakness, or frailty in older women. Experimental Gerontology. 2005;40(4):344–352. [PubMed: 15820616]
  144. Walston J, Xue Q, Semba RD, Ferrucci L, Cappola AR, Ricks M, Guralnik J, Fried LP. Serum antioxidants, inflammation, and total mortality in older women. American Journal of Epidemiology. 2006;163(1):18–26. [PubMed: 16306311]
  145. Weaver J, Huang MH, Albert M, Harris T, Rowe J, Seeman TE. Interleukin-6 as a predictor of cognitive function and cognitive decline: MacArthur Studies of Successful Aging. Neurology. 2002;59:371–378. [PubMed: 12177370]
  146. Wen M, Hawkley LC, Cacioppo JT. Objective and perceived neighborhood environment, individual SES and psychosocial factors, and self-rated health: An analysis of older adults in Cook County, Illinois. Social Science Medicine. 2006;63:2575–2590. [PubMed: 16905230]
  147. Wilson RS, Bienias JL, Evans DA, et al. The Religious Orders Study: Overview and relation between change in cognitive and motor speed. Aging Neuropsychology and Cognition. 2004;11:280–303.
  148. Yende S, Waterer GW, Tolley EA, Newman AB, Bauer DC, Taaffe DR, Jensen R, Crapo R, Rubin S, Nevitt M, Simonsick EM, Satterfield S, Harris T, Kritchevsky SB. Inflammatory markers are associated with ventilatory limitation and muscle dysfunction in obstructive lung disease in well functioning elderly subjects. Thorax. 2006;61(1):10–16. [PMC free article: PMC2080698] [PubMed: 16284220]
  149. Zhang HP, Singer B. Recursive partitioning in the health sciences. New York: Springer-Verlag; 1999.