Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Psychosom Med. Author manuscript; available in PMC 2011 January 1.
Published in final edited form as:
PMCID: PMC2807361

Subjective Socioeconomic Status and Presence of the Metabolic Syndrome in Midlife Community Volunteers



Objective indices of socioeconomic status (SES) predict diverse sources of morbidity and mortality, as well as numerous biological and behavioral risk factors for disease. Here we examine whether subjective SES may be similarly associated with measured risk factors, including the metabolic syndrome and its components of elevated blood pressure, high fasting glucose, dyslipidemia, and central adiposity.


Observations were based on a community sample of 981 adults (30–54 years of age; 52% female; 84% white, 16% African American). Subjective SES was measured using the nationally referenced (U.S.) MacArthur Scale of Subjective Social Status, and objective SES indexed by composite of years of education and family income.


Likelihood of meeting criteria for presence of the metabolic syndrome varied inversely with subjective SES (Odds Ratio [OR] =0.75; 95% CI: 0.64, 0.88, for a 1 SD increase in subjective SES, adjusted for age, sex, and race), and this association persisted on further adjustment for objective SES (OR = 0.82; 95% CI: 0.68, 0.99). Subjective SES was also associated inversely with blood pressure, waist circumference, and serum triglycerides, and positively with HDL cholesterol. Level of physical activity and smoking status were predicted by subjective SES as well, but adjusting for these health behaviors did not appreciably reduce associations of subjective SES with metabolic syndrome and syndrome components.


These findings support speculation that perceived social standing is associated with prominent cardiovascular risk factors and may prove a useful adjunct to conventional socioeconomic indicators in epidemiological research.

Keywords: socioeconomic status, subjective social status, metabolic syndrome, blood pressure, central adiposity, dyslipidemia


Inequalities in income, education, and other socioeconomic indicators predict diverse sources of morbidity and mortality, as well as numerous biological and behavioral risk factors for disease (13). These associations span multiple gradations of socioeconomic status (SES) and are not explained by poverty alone, illiteracy, or restricted access to health care resources (2). That health outcomes may vary with socioeconomic inequalities even among those who are not materially disadvantaged has prompted speculation that subjective experiences of relative social standing may partly account for SES-related variability in disease risk (4, 5). Supporting such speculation, people who report they occupy a lower social position than others on a visual ladder depicting ordered rungs of ascending SES also report poorer health (611). Moreover, several studies have shown that reports of lower subjective SES are associated with dysregulated adrenocortical responses to stress, poorer sleep latency, higher heart rate, and greater adiposity and body mass (613). It is noteworthy that many of these associations persist after multivariate adjustment for objective socioeconomic indicators, suggesting that measures of subjective SES may capture properties of positional status that correlate imperfectly with conventional SES indicators (e.g., educational attainment, income, occupational grade), yet also predict health status or markers of disease risk (613). Owing to reliance on measures of self-rated health in much of this literature, however, it is possible that apparent associations of subjective social status with experienced health are inflated by reporting biases or other attributes of shared variance common to self-report data (8). Accordingly, in this study we tested whether subjective SES is associated with measured (rather than self-rated) health risk, as indicated by presence of the metabolic syndrome and its component factors of elevated fasting glucose and blood pressure, dyslipidemia, and central adiposity, in a community sample of mid-life adults.


Study Participants

Data were derived from the University of Pittsburgh Adult Health and Behavior (AHAB) registry, which is comprised of behavioral and biological measurements obtained on non-Hispanic white and African American individuals recruited between 2001 and 2005 via mass-mail solicitation from communities of southwestern Pennsylvania, USA (principally Allegheny County) (Cf., 14 – 16). Subjects were 1046 male (49%) and female AHAB participants, aged 30–54 years. Eighty-four percent of participants were White and 16% African American. Study exclusions were a reported history of atherosclerotic cardiovascular disease, chronic kidney or liver disease, cancer treatment in the preceding year, and major neurological disorders, schizophrenia or other psychotic illness. Other exclusions included pregnancy and the use of insulin, glucocorticoid, antiarrhythmic, psychotropic, or prescription weight-loss medications. Data collection occurred over multiple laboratory sessions, and informed consent was obtained in accordance with approved protocol and guidelines of the University of Pittsburgh Institutional Review Board.

Metabolic Syndrome and Risk Factor Assessments

Components of the metabolic syndrome were assessed in the morning following a 12-hr, overnight fast. Blood pressure measurements were obtained by trained staff, using a mercury sphygmomanometer and cuff size appropriate to the participant’s arm circumference. Resting blood pressure was measured from the right arm and calculated as the mean of two consecutive readings obtained in a seated position. Measurements were obtained of participants’ height, weight, and waist circumference. Determination of fasting serum lipids, glucose, and insulin was performed by the Heinz Nutrition Laboratory, University of Pittsburgh Graduate School of Public Health, as described previously (17). The metabolic syndrome was defined by current American Heart Association/National Heart, Lung, and Blood Institute criteria (18,19) as the presence of three or more of the following: 1) waist circumference ≥102 cm in men or ≥88 cm in women; 2) fasting serum glucose ≥100 mg/dl or use of oral hypoglycemic medication; 2) blood pressure ≥130 mmHg systolic/85 mmHg diastolic or use of antihypertensive medication; 4) serum triglycerides ≥150 mg/dl or medication for hypertriglyceridemia; and 5) HDL cholesterol <40 mg/dl in men or <50 mg/dl in women, or use of medication for low HDL cholesterol. In addition, insulin resistance was estimated by homeostatic model assessment, calculated as follows: insulin resistance (HOMA-IR) = serum insulin (µIU/ml) -x- fasting blood glucose (mmol/liter)/22.5 (20).

Nineteen subjects were excluded from analysis due to missing data on one or more components of the metabolic syndrome. Also excluded were 45 participants taking cholesterol-lowering drugs (e.g., statins), because use of these medications renders triglyceride and HDL cholesterol (and therefore, metabolic syndrome) status indeterminate. The final sample included 981 subjects (474 men; 507 women). The 19 participants (2%) excluded for missing data did not differ from the final sample in age, distribution by sex or race, or subjective SES, but had about two years’ less education (t998 = −2.9, p < .01) and lower annual income (t995 = −3.6, p < .001). In contrast, the 45 subjects (4%) excluded for use of cholesterol-lowering medications were equivalent to the final sample on all socioeconomic indicators, but consistent with risk factors for hypercholesterolemia, were older by 4 years (t1025 = 4.4, p < .001) and more likely to be male (65% vs. 48%; χ2 = 4.4, p = .04).

Socioeconomic Status

Subjective SES was assessed using the nationally referenced MacArthur Scale of Subjective Social Status (6; This scale depicts relative social standing as the ten rungs of a pictured ‘social’ ladder. Respondents are asked to place an ‘X’ on the rung that best indicates where they stand in relation to others in the United States population by reference to income, education, and occupational prestige. (See figure, Supplemental Digital Content 1, for depiction of scale). Objective SES was indexed by two conventional indicators: 1) cumulative years of schooling; and 2) annual (pre-tax) family income, within bracketed ranges of <$25,000, 25–34,999; 35–49,999, 50–64,999, 65–80,000, and >$80,000. As in prior reports (21, 22), we computed a composite measure of objective SES by averaging the standardized (z-score) values of the two index variables for each individual. This measure was then re-standardized to yield of a distribution with mean of 0.0 and SD of 1.0.

Health Behaviors

Perceived social standing has been shown previously to predict cigarette smoking and sedentary lifestyle, and both are known to increase cardiovascular risk. To determine if any association of subjective SES with the metabolic syndrome or its components might be accounted for by correlated variation in these health behaviors, we assessed smoking status by self-report (current vs. previous/never) and level of physical activity by the Paffenbarger Physical Activity Questionnaire (23). The latter instrument estimates kilocalorie expenditure (kcal) over a 7-day period (e.g., blocks walked, stairs climbed, leisure time activity) and has good predictive validity for coronary heart disease (24).

Statistical Analysis

Association of subjective and objective SES indicators with presence of the metabolic syndrome was first evaluated by simple logistic regression without covariates. In these analyses individually entered predictors were subjective SES (i.e., the MacArthur Scale of Subjective Social Status), objective SES, and standardized values of the two components of objective SES, education and income. We then further evaluated SES associations in two hierarchical logistic regression models in which age (years), sex (coded 0 [M], 1 [F]), and race (coded 0 [White], 1 [African American]) were entered first (Step 1) due to prior evidence that presence of the metabolic syndrome or components of the syndrome vary as a function these demographic factors (25). In Model I, subjective SES was then entered on Step 2. In Model II, we entered objective SES (‘income/education’) after covariates (Step 2) and subjective SES on Step 3, to determine if social ladder scores predicted presence of the metabolic syndrome above and beyond objective SES. This sequence of analyses was next repeated for each of the components of the metabolic syndrome (i.e., presence/absence of elevated blood pressure, central adiposity, high fasting glucose, high triglycerides, and low HDL cholesterol concentration). Estimated odds ratios (ORs) derived from logistic regressions are referenced to a one SD increase in the standardized subjective and objective SES indicator variables. Finally, parallel analyses were conducted by hierarchical linear regression on continuously distributed metabolic syndrome components, excluding individuals taking prescribed antihypertensive, oral hypoglycemic, or dyslipidemic medications at the time of study participation. Dependent variables in these analyses were systolic and diastolic blood pressure, waist circumference, fasting glucose concentration, triglycerides and HDL cholesterol, as well as fasting insulin and HOMA-IR. Linearity of association was examined in analyses of continuously distributed dependent variables by addition of non-linear (quadradic, cubic) terms to regression models, but in no instance did these qualify interpretations of linear trend. Also, prior to analysis, a natural log (ln) transformation was applied to fasting glucose, insulin and triglyceride concentration to better approximate normal distribution.

Finally, to explore the potential contribution of health behaviors to associations between SES and metabolic risk factors, we repeated the foregoing analyses, but entered smoking status (coded 1 [current] or 0 [previous/never]) and physical activity ([kcal/wk]/1000) in Step 2 of each regression model (i.e., after demographic covariates). Then, in Model III, subjective SES was entered on Step 3 to determine if social ladder scores predicted presence of the metabolic syndrome, as well as criterion elevations and continuous variation in syndrome components, independently of these health behaviors. And last, In Model IV, we entered objective SES on Step 3, followed by subjective SES in Step 4.


Subjects included in the present analyses reported a mean subjective SES score of 6.2 ± 1.6, with the following proportion of all participants marking each of the ten ladder rungs (from lowest to highest subjective SES): 0.3, 2, 4, 8, 15, 27, 26, 12, 5, and 0.7%. Participants averaged 15.5 ±2.6 years of schooling (range: 8–24 years), with a broad distribution of educational attainments (without High School diploma [3%]; completed High School only [115%]; post-secondary education without Bachelors degree [29%]; Bachelors degree [38%]; Masters or equivalent professional degree [15%]; Doctoral or doctoral-level professional degree [4%]. The proportion of subjects falling within the six ranges of reported income (lowest to highest) were 18.0, 11.0, 16.0, 17.0, 15.0, and 23.0%. Across subjects, subjective SES (ladder rankings) correlated significantly with years of education (r = 0.38, p < .001), income range (r = 0.52, p < .001), and our composite index of objective SES (r = 0.57, p< .001), indicating that individuals earning a lower income and completing fewer years of schooling tended also to perceive themselves as holding a lower social standing than more “objectively” advantaged study participants. Nonetheless, variability in subjective SES scores could be accounted for only partially by overlapping variation in the composite measure of objective SES derived from income and education.

On average, the sample was slightly overweight and had a mildly elevated mean fasting glucose concentration (see Table 1). Average blood pressure, and triglyceride and HDL concentrations fell within normal limits. Prevalence of the metabolic syndrome was 24%, and criterion values for the presence of syndrome components were met at the following frequencies: elevated blood pressure (34%), high triglycerides (25%), high fasting glucose (31%), high waist circumference (35%), and low HDL cholesterol (27%). With respect to health behaviors, 18% of the sample smoked cigarettes at the time of study participation, and 21% may be considered sedentary by expenditure of <1000 kcal/wk in physical activity (26)

Characteristics of Subjects Included in Analyses of the Metabolic Syndrome

Metabolic Syndrome

As shown in Table 2, higher subjective SES was associated with significantly lower odds of meeting criteria for presence of the metabolic syndrome. Thus, individuals who reported occupying a higher social rung on the MacArthur ladder were found less likely to have the metabolic syndrome than those placing themselves on a lower rung. The same was seen for our index of objective SES and each of its two components, although the former was a somewhat stronger predictor than years of education or level of reported income separately. When age, sex and race were entered simultaneously into logistic regression (see Table 3, Step 1), likelihood of exhibiting the metabolic syndrome rose with age and was substantially lower in women than men, but did not differ between Whites and African Americans. When entered next (in Model I), subjective SES again predicted presence of the metabolic syndrome. Here, an increase of one SD in the sample distribution of subjective SES scores (equal to about 1.6 rungs “up” on the MacArthur ladder) was associated with a 25% reduction in odds of meeting full metabolic syndrome criteria. In Model II, objective SES again predicted presence of the metabolic syndrome and did so at about the same strength of association as subjective SES. Inclusion of interaction terms showed neither subjective nor objective SES to vary in their association with the metabolic syndrome as a function of age, sex, or race (p’s > 0.05). Finally, when subjective SES was entered on Step 3 (i.e., after both demographic covariates and objective SES), subjective SES continued to predict presence of the metabolic syndrome, albeit more weakly (OR = 0.82).

Socioeconomic Indicators (Subjective and Objective SES, Years of Education and Income) as Predictors of NCEP-defined Metabolic Syndrome and Syndrome Components. (N=981). (Distributions of all socioeconomic predictors are standardized to a mean of 0.0 ...
Demographic and Socioeconomic Predictors of NCEP-defined Metabolic Syndrome and Syndrome Components. (N981).

Components of the Metabolic Syndrome

With respect to syndrome components, unadjusted analyses (Table 2) showed subjective SES to predict criterion elevations in blood pressure, high waist circumference, and low HDL cholesterol. Objective SES did so similarly, in addition to predicting high fasting glucose, whereas education and income alone were less consistently associated with individual risk factors.

As shown in Table 3, age and sex predicted criterion elevations in blood pressure, fasting glucose and triglyceride values, but neither variable predicted sex-specific thresholds for low HDL cholesterol or central adiposity. Relative to Whites, African American participants were twice as likely to meet criteria for elevated blood pressure and high waist circumference and were about 50 percent more likely to have a high fasting glucose concentration. However, neither HDL cholesterol nor presence of the metabolic syndrome varied by race, and African Americans were significantly less likely to have elevated serum triglycerides. When SES indicators were next tested individually after entering demographic covariates (i.e., in Step 2, Models I and II), both subjective and objective SES predicted criterion elevations in blood pressure, waist circumference, and triglycerides, and low HDL cholesterol. Further, subjective SES continued to predict an elevated blood pressure and high waist circumference when entered into logistic models after objective SES, but corresponding associations were not significant for triglycerides or HDL cholesterol. Only objective SES was associated with high fasting glucose, but it may be noted that this relationship loses significance if the six diabetic participants taking oral hypoglycemic medication are removed from the analysis (OR = 0.83; 95% CI: 0.66, 1.04; p = .11).

As a component of the metabolic syndrome, elevated blood pressure can be identified by any of several criteria, including current antihypertensive treatment and elevations in either systolic or diastolic blood pressure, or both. In the present sample, 21% of subjects meeting syndrome criteria for elevated blood pressure (69 of 332 participants) were prescribed antihypertensive medications. As treated hypertensives, the blood pressures of these participants (if unmedicated) would be expected to predominate in the upper ranges of blood pressure for the study sample. Also, because levels of blood pressure needed to meet criteria for inclusion in the metabolic syndrome are much lower than clinical thresholds for hypertension diagnosis or antihypertensive treatment, untreated subjects with a syndrome-defined blood pressure elevation will include many subjects whose blood pressures hover close to threshold values. In fact, 39% of untreated participants who met syndrome criteria for an elevated blood pressure here did so with systolic and/or diastolic pressures within the narrow range of 130–139/85–89 mm Hg and 84% had blood pressures <150/95 mm Hg.

Owing to these considerations, we asked whether socioeconomic indicators would predict blood pressure elevations even within the compressed range of “higher” values represented in the sample when treated hypertensives were excluded. After adjusting for demographic covariates, subjective SES did predict threshold elevations in measured blood pressure (OR = 0.79, 95% CI: 0.68, 0.93, p < .01), but objective SES did not (OR = 0.88, 95% CI: 0.75, 1.03; p = .12). Interestingly, tests for interactions with demographic variables showed participant age to moderate the association of subjective SES with an elevated blood pressure (p = .04). Ladder scores strongly predicted criterion elevations among subjects above the median age of 46 yr (OR = 0.65; 95% CI: 0.52, 0.82; p < .001), but were unrelated among younger participants (OR = 0.99; 95% CI: 0,80, 1.23); p = .93). Age did not interact similarly with objective SES, and no similar interaction was observed on analyses of other syndrome components or the metabolic syndrome itself. In sum, objectively defined socioeconomic variation predicted an elevated blood pressure in this sample only when treated hypertensives were included in analysis, whereas subjective SES did so also when treated subjects were excluded, but then only among older participants.

Syndrome components as continuous variables

Linear regressions predicting variation in systolic and diastolic blood pressure, waist circumference, fasting glucose concentration, triglycerides and HDL cholesterol are summarized in Table 4. The patterns of association with covariates were virtually the same as for dichotomized syndrome components (in Table 3), except that here: 1) age predicted HDL cholesterol; and 2) compared to women, men had higher waist circumference and lower HDL cholesterol. The latter differences simply reflect the fact that inclusion criteria for metabolic syndrome specify sex-specific thresholds for waist circumference and HDL cholesterol.

Demographic and Socioeconomic Predictors of Blood Pressure (unmedicated, N=912), Waist Circumference (N=981), and Fasting Glucose (unmedicated, N=975), Triglycerides and HDL Cholesterol (unmedicated, N=973).

In contrast to the analysis of dichotomized blood pressure elevation (defined by systolic pressure ≥130 and/or diastolic pressure ≥85 mmHg, or antihypertensive treatment), neither subjective nor objective SES straightforwardly predicted variation in systolic or diastolic blood pressure in analyses restricted to the 912 untreated study participants. This is consistent with the failure of objective SES to predict threshold elevations of blood pressure in analyses excluding treated hypertensives. And like our previous analysis of dichotomized blood pressure elevations, participant age was found to moderate an association of subjective (but not objective) SES with measured blood pressure. The interaction of subjective SES with age was significant for both systolic and diastolic blood pressure (p.s ≤.01), and as in the earlier analysis of dichotomized variables, did not extend to continuous variation in any other syndrome component. Among participants above the median age, higher ladder rankings predicted lower systolic (b = −1.48, se = 0.64, β = −0.11, p =.02) and diastolic blood pressure (b = −1.11, se = 0.41, β = −0.12, p <.01), and the latter association persisted after adjustment for objective SES (b = −1.00, se = 0.48, β = −.11, p =.04). Among younger subjects, no similar association was observed for either systolic (b = 0.89, se = 0.64, β = .07, p =.17) or diastolic blood pressure (b = 0.08, se = 0.42, β = 0.01, p = 85).

With respect to other syndrome components, both SES indices accounted for significant variance in serum triglycerides and waist circumference, and subjective SES predicted quantitative variation in the latter risk factor after adjustment for objective SES. Only subjective SES was associated with HDL cholesterol, and neither subjective nor objective SES accounted for variation in fasting glucose concentration. Because measures of fasting insulin and HOMA-IR were available in this sample as well, we examined these variables in parallel regression models. Unlike the findings for glucose, subjective SES was a significant predictor of both insulin (log transformed, b = −0.05, se = 0.02, β= −0.09, p < .01) and HOMA-IR (b = −0.17, se = 0.08, β= −0.07, p < .04) after adjustment for covariates, as was objective SES (Insulin: b = −0.95, se = 0.26, β = −0.12, p< .001; HOMA-IR: b = −0.29, se = 0.08, β = −0.12, p < .001). Finally, when entered after objective SES, subjective SES did not account for significant additional variation in insulin or HOMA-IR.

Health Behaviors

In this sample, women were less likely to smoke than men (OR = 0.49; 95% CI: 0.35, 0.69; p < .001), while Black participants were more likely than Whites (OR = 2.88; 95% CI: 1.97, 4.22; p < .001). Men engaged in more physical activity than women (2602 [se = 88] vs 2321 [79] kcal/wk; t979 = 2.4, p = .02), as did Whites relative to Blacks (2560 [64] vs 1908 [142] kcal/wk; t979 = 4.1, p < .001). Age was unrelated to both smoking (OR = 0.99; 95% CI: 0.96, 1.01); p = .37) and physical activity (r980 = −0.05, p = .13). After adjustment for demographic covariates, smoking status was predicted by both subjective (OR = 0.72; 95% CI: 0.61, 0.85; p < .001) and objective SES (OR = 0.57; 95% CI: 0.47, 0.68; p < .001). Likewise, higher subjective SES was associated with a higher level of physical activity (b = 0.24, se = 0.06, β = .13, p < .001), as was objective SES (b = .17, se = .06, β = .09, p <.01). When entered after objective SES, ladder rankings continued to predict activity level (b = 0.22, se = 0.07, β = .12, p <.01), but were not independently associated with smoking status (OR = 0.94; 95% CI: 0.76, 1.14; p = .52).

The relationship of these two health behaviors to presence of the metabolic syndrome and syndrome components, after adjusting for age, sex and race, is presented in Table 5 (for dichotomized risk indices) and Table 6 (continuous variables). Smoking was associated with an elevated fasting glucose, but not with the metabolic syndrome itself or threshold elevations in blood pressure, waist circumference and triglycerides, or low HDL cholesterol. However, smoking did predict variation in systolic blood pressure, central adiposity, and fasting glucose concentrations when these were analyzed as continuous variables. Higher levels of physical activity was associated with a reduced likelihood of meeting criteria for the metabolic syndrome, as well as four of the five syndrome components, and predicted levels of all continuously distributed risk factors. Smoking was associated with higher fasting insulin concentrations (log transformed, b = −.10, se = 0.04, β = −.08, p = .02), and physical activity predicted both fasting insulin (b = −.04, se = 0.01, β = −.15, p < .001) and HOMA-IR (b = −.19, se = .04, β = −.14, p < .001).

Smoking Status, Physical Activity, and Socioeconomic Predictors of NCEP-defined Metabolic Syndrome and Syndrome Components (N=980).
Smoking Status, Physical Activity, and Socioeconomic Predictors of Blood Pressure (unmedicated, N=912), Waist Circumference (N=981), and Fasting Glucose (unmedicated, N=975), Triglycerides and HDL Cholesterol (unmedicated, N=973).

Metabolic syndrome and syndrome components

After controlling for age, sex, race, smoking status and physical activity, both subjective and objective SES showed the same pattern of associations with the metabolic syndrome and with syndrome components (either as dichotomized or continuous variables) seen in previous analyses adjusting for demographic covariates alone, although the strength of these relationships (ORs and regression coefficients) was uniformly, if slightly, weaker (Table 5 and Table 6). In fully adjusted models, fasting insulin was predicted by both subjective (log transformed, b = −.04, se = 0.02, β = −.08, p = .01) and objective SES (b = −.06, se = 0.02, β = −.13, p < 001); HOMA-IR was similarly predicted by objective SES (b = −.30, se = 0.08, β = −.12, p <.001), but was only marginally associated with subjective SES (b = −.15, se = 0.08, β = .06, p = .07).

Although socioeconomic indicators did not predict variation in blood pressure after adjustment for demographic covariates and health behaviors in analyses excluding treated hypertensives (Table 6, Models III and IV), tests for age-dependent interactions again showed participant age to moderate an association of subjective SES with both systolic and diastolic measurements (p’s < .01). Among subjects above the median age, ladder rankings covaried inversely with systolic (b = −1.29, se = 0.65, β = −0.09, p = .049) and diastolic blood pressure (b = −0.99, se = 0.42, β = −.11, p = .02) after controlling for smoking and physical activity, and in a fully adjusted model including objective SES, ladder scores continued to predict diastolic blood pressure, albeit only marginally (b = −0.90, se = 0.48, β = −0.10, p = .06). Corresponding analyses among younger subjects showed subjective SES unrelated to systolic (b = 0.46, se = 0.55, β = 0.04, p = .40) or diastolic blood pressure (b = 0.00, se = 0.43, β = 0.00, p = .99).


Subjective SES -- assessed by a social ladder measure of relative social standing – has emerged as a novel indicator of social inequalities relevant to health, although its validation rests largely on studies of self-reported health and illness (611) or non-standard risk factors (e.g., sleep latency, neuroendocrine responses to stress) (6). Here, we extend prior work on subjective SES to encompass previously less-studied, but common indices of measured cardiovascular risk among mid-life adults. Individuals who placed themselves on the higher rungs of a visual ladder depicting relative social standing were less likely than those reporting a lower standing to meet criteria for the metabolic syndrome, a constellation of abnormalities that confers heightened risk for diabetes, incident cardiovascular events, and cardiovascular and all-cause mortality (27, 28). Moreover, subjective SES appeared to be as strongly related to the metabolic syndrome as a simple survey-based objective SES measurement, determined here by a composite of income and education. Expressed in units of the ladder of subjective SES, for instance, two rungs “up” on the social ladder (which is equivalent to the interquartiile range for this measure) corresponded to a 30% reduction in estimated odds of meeting full syndrome criteria (i.e., presence of three or more risk factors). Although modest, this association is equal to the protective benefit predicted for an increase of 1400 kcal/wk in physical activity. Conversely, the increased risk reflected in a descent of two rungs on the social ladder is equivalent to that associated with about 8 years of aging across the mid-life range represented in this sample.

In prior work, similarly assessed subjective SES covaried inversely with adolescent obesity and central adiposity (waist-hip ratio) in middle-aged women (6, 12, 13). Individual differences in waist circumference and, as a component of the metabolic syndrome, “high” waist circumference were likewise associated here with lower subjective SES, as well as lower objective SES. Both SES indices also predicted “low” HDL cholesterol concentration and threshold elevations in serum triglycerides and blood pressure. In contrast to the dichotomous classification of blood pressure, though, which is based on the presence or absence of threshold elevations in systolic and/or diastolic measurements or antihypertensive treatment, neither subjective nor objective SES showed a straightforward association with systolic or diastolic blood pressure when these were analyzed separately as continuous variables among untreated subjects. However, participant age moderated a relation of subjective SES with both systolic and diastolic blood pressure, such that higher ladder rankings were associated with lower blood pressures among older, but not younger, subjects. This age-dependent relationship was not seen for other syndrome components, where (except for the association of objective SES with HDL cholesterol) results were generally similar when risk factors were examined as dichotomous or continuously distributed variables.

Because most models of the metabolic syndrome posit insulin resistance as a key factor in the aggregation of component risk factors, it is perhaps surprising that we did not also find elevated fasting glucose predicted by low subjective SES. On the other hand, both subjective and objective SES covaried inversely with fasting insulin and estimated insulin resistance (HOMA-IR). It may be noted that early in the development of insulin resistance, the pancreas compensates by secreting more insulin, and this hyperinsulinemia helps maintain euglycemia (29). Thus, in relatively young cohorts with a low prevalence of diabetes such as ours (mean age, 45 years; 3% diabetic), elevated fasting insulin and other indices of insulin resistance may be more sensitive markers of disrupted glucose metabolism than fasting hyperglycemia (30).

It is noteworthy that associations of subjective SES with the cardiovascular risk factors studied here occasionally persisted after controlling for objective SES (e.g., the dichotomized threshold for blood pressure elevation, diastolic blood pressure among older participants, waist circumference, and HDL cholesterol). Albeit with some inconsistency, similarly independent effects of subjective SES have been reported previously for measures of self-rated health, heart rate, BMI, adolescent obesity, and blood pressure (69, 12, 13, 31). Together, these findings suggest that perceived social standing may tap aspects of social stratification that elude detection by more conventional socioeconomic metrics, but which likewise account for variation in cardiovascular disease risk factors and self-reported health. However, two alternative hypotheses may be suggested. First, it is possible that certain correlates of health status, such as obesity, might influence self-perceptions of social standing (reverse causation). And second, it is conceivable that ratings of subjective SES reflect reporting biases relating to subject characteristics (e.g., personality) that are themselves associated with indices of health and disease risk.

Regarding the first of these possibilities, social stigma stemming from cultural norms defining a desirable body weight and fat distribution might lead some obese individuals to report lower subjective social status, even though subjects are instructed to reference ladder rankings on the MacArthur scale to dimensions of income, education, and occupational prestige. If data of the present study are re-analyzed after excluding the 27% of participants who were obese (viz., subjects with BMI ≥30), however, subjective SES continues to predict presence of the metabolic syndrome (OR = 0.78; 95% CI: 0.62, 0.98; p = .04). And if BMI is instead entered as an additional covariate on analysis of all subjects, subjective SES is again similarly associated with the metabolic syndrome (OR = 0.81; 95% CI: 0.67, 0.99; p = .03). We nonetheless recognize that hypotheses of reverse causation are difficult to fully address in cross-sectional analyses and that alternative explanations may remain.

With respect to non-socioeconomic determinants of subjective SES ratings, Singh-Manoux et al. (2003) have shown that individuals do primarily reference socioeconomic parameters in assigning themselves to a rung on the MacArthur ladder, after which potentially biasing personality characteristics (such as traits of hopelessness, optimism, or hostility) account for little additional variance. Hence, the authors suggest that subjective SES may convey a “cognitive averaging” of multiple dimensions of socioeconomic circumstance – e.g., earnings and accumulated wealth (or financial security), prestige accorded certain educational attainments, occupational grade, or standard of living, and beyond personal attributes, parental SES or area-level indices, such as neighborhood quality and community resources -- only some of which may be captured by the few “objective” markers of SES usually collected in individual epidemiologic studies. It is also likely, however, that the proportion of variance in subjective SES unaccounted for by conventional socioeconomic indicators is not fixed, but varies between cultures or geographic regions. For instance, while ladder rankings correlated only moderately with years of schooling (r = .38) and income (r = .52) in the present sample, a prior study of mortality predictors among middle-aged Hungarian men and women found a much closer correspondence between subjective SES and both education (men: r = .82; women: r = .50) and income (men: r = .83; women: r = .66) (32).

Whatever the relationship between indices of subjective and objective SES, it is important to consider how socioeconomic factors might affect cardiovascular risk. One obvious mechanism is through health impairing attributes of habit and lifestyle. In our sample, current smoking status and level of physical activity were predicted by both ladder rankings and objective SES, which is consistent with prior literature (33, 34). And while smoking was associated with lower blood pressure and waist circumference (but higher fasting glucose and insulin concentrations), lower physical activity predicted presence of the metabolic syndrome and virtually all syndrome components, as well as fasting insulin and estimated insulin resistance (HOMA-IR). Nonetheless, adjustment for these health behaviors only slightly attenuated the relationship of subjective or objective SES with the metabolic syndrome and associated risk factors. Of course, other lifestyle factors, such as dietary preferences, might still account in part for associations between socioeconomic indicators and cardiovascular risk (31).

Another frequently cited correlate of low SES is the stress (or distress) that may be occasioned by heightened challenges of daily living, uncertainties of future prospect, or demoralization stemming from perceptions of relative social or material disadvantage (5, 3537). In turn, putative stress-related dimensions of social stratification might be gauged more sensitively by subjective estimate than by conventional measures of objective SES. Lower rankings on the MacArthur social ladder, for instance, have been associated with impaired sleep and elevated heart rate (6), with an exaggerated rise in the stress-hormone, cortisol, on awakening from sleep (38), and with a non-habituating cortisol response to acute psychological stress (6). In most instances these associations were independent of objective SES indicators. Recently, too, lower subjective SES correlated with reduced gray matter volume in the perigenual area of the anterior cingulate cortex (ACC), a brain region involved in emotional experience and the regulation of behavioral and physiological reactivity to stress (35). Volumetric changes in the ACC and other paralimbic brain areas have been documented in stress-related psychiatric symptomatology, such as depression (e.g., 39), which is also predicted by low subjective status (8, 12). Of potential relevance to the present findings, this neurobiologic correlate of subjective SES may modulate key circuitries of autonomic and neuroendocrine control affecting cardiovascular and metabolic regulation. For example, reduced ACC volume has been associated with a dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis and may therefore account for the altered cortisol responsivity seen in individuals of low subjective social status (6, 38). Importantly, HPA dysregulation is related also to several components of cardiovascular risk, including obesity, insulin resistance, dyslipidemia, elevated blood pressure, preclinical atherosclerosis, and the metabolic syndrome (4043). It is possible, then, that a portion of the contribution of subjective SES to disease risk derives from psychophysiologic correlates of perceived social standing, particularly those associated with accompanying experiences of psychological stress.

Even though we observed significant associations of subjective and objective socioeconomic indices with metabolic syndrome and its components, interpretation of these findings is qualified by several study limitations. The sample is skewed toward the higher end of educational attainment, for instance, with the majority of subjects having completed college and very few lacking at least a secondary level of education. The range of participant ages included in the AHAB registry is restricted to mid-life adults (30–54 years), and ethnic heterogeneity of the sample is limited (84% white, of non-Hispanic origin; 16% African American). Exclusions for prevalent disease and several categories of prescription medication also define the study sample as generally healthy, and all participants were recruited from a common geographic location (southwestern Pennsylvania). While these factors preclude generalization of study findings to other populations, it is also possible that the associations seen here would be magnified in samples hosting a greater variability of socioeconomic circumstance, greater ethnic diversity, and expanded variation in the health status of study participants. Nonetheless, prevalence of the metabolic syndrome in the current sample (24%) was not dissimilar to that reported in population cohorts (25, 44), and within the constraints of our sample, associations of socioeconomic indices with presence of the metabolic syndrome did not vary by sex or race. Due to the cross-sectional nature of these analyses, though, the direction of associations between dimensions of SES and the metabolic syndrome or its component risk factors (and by implication, any causal influence of subjective or objective SES on these outcomes) is uncertain. Hence, future extension of these observations to fully prospective investigation is warranted.

Summary and Implications

In this study, a single-item measure of perceived social standing – subjective SES – was found to predict presence of the metabolic syndrome and associated indices of measured cardiovascular risk, including elevated blood pressure, central adiposity, high triglyceride and low HDL cholesterol concentrations, fasting insulin levels and estimated insulin resistance. While subjective SES also predicted smoking status and reported levels of physical activity, its association with the metabolic syndrome and syndrome components persisted on adjustment for these two health-related behaviors. Finally, effects of subjective SES were similar to those associated with an objective SES index (composite of income and education), and in some instances were independent of correlated variation in objective SES. In sum, our findings support speculation that perceived social status is associated with prominent cardiovascular risk factors and many prove a useful adjunct to conventional socioeconomic indicators in epidemiological research.

Supplementary Material



Grant acknowledgment: This research was supported by NIH grants PO1 HL040962 and RO1 HL065137 to SBM.


socioeconomic status


Supplemental Digital Content List

Supplemental Digital Content 1. Figure depicting MacArthur Scale of Subjective Social Status. doc

Contributor Information

Stephen B. Manuck, University of Pittsburgh, Pittsburgh, PA, Department of Psychology.

Jennifer Phillips, University of Pittsburgh, Pittsburgh, PA, Department of Psychology.

Peter J. Gianaros, University of Pittsburgh, Pittsburgh, PA, Department of Psychiatry.

Janine D. Flory, Queens College, City University of New York, Department of Psychology.

Matthew F. Muldoon, University of Pittsburgh, Pittsburgh, PA, Division of Clinical Pharmacology.


1. Davey Smith G, Bartley M, Blane D. The Black Report on socioeconomic inequalities 10 years on. Br Med J. 1990;301:373–377. [PMC free article] [PubMed]
2. Adler NE, Boyce T, Chesney MA, Cohen S, Folkman S, Kahn R, Syme SL. Socioeconomic Status and Health: The Challenge of the Gradient. Am Psychol. 1994;49:15–24. [PubMed]
3. Backlund E, Sorlie PD, Johnson NJ. The Shape of the Relationship Between Income and Mortality in the United States: Evidence from the National Longitudinal Mortality Study. Ann Epidemiol. 1996;6:12–20. [PubMed]
4. Marmot M, Wilkinson RG. Social determinants of health. Oxford: Oxford University Press; 1999.
5. Wilkinson RG. Health, hierarchy, and social anxiety. Ann N Y Acad Sci. 1999;896:48–63. [PubMed]
6. Adler NE, Epel ES, Castellazzo G, Ickovics JR. Relationship of Subjective and Objective Social Status with Psychological and Physiological Functioning: Preliminary Data in Healthy White Women. Health Psychol. 2000;19:586–592. [PubMed]
7. Ostrove JM, Adler NE, Kuppermann M, Washington AE. Objective and subjective assessments of socioeconomic status and their relationship to self-rated health in an ethnically diverse sample of pregnant women. Health Psychol. 2000;19:613–618. [PubMed]
8. Singh-Manoux A, Adler NE, Marmot MG. Subjective social status: its determinants and its association with measures of ill-health in the Whitehall II study. Soc Sci Med. 2003;56:1321–1333. [PubMed]
9. Singh-Manoux A, Marmot MG, Adler NE. Does subjective social status predict health and change in health status better than objective status? Psychosom Med. 2005;67:855–861. [PubMed]
10. Hu P, Adler NE, Goldman N, Weinstein M, Seeman TE. Relationship between subjective social status and measures of health in older Taiwanese persons. J Am Geriatr Soc. 2005;53:483–488. [PubMed]
11. Franzini L, Fernandez-Esquer ME. The association of subjective social status and health in low-income Mexican-origin individuals in Texas. Soc Sci Med. 2006;63:788–804. [PubMed]
12. Goodman E, Adler NE, Kawachi I, Frazier AL, Huang B, Colditz GA. Adolescents' perceptions of social status: development and evaluation of a new indicator. Pediatrics. 2001;108:1–8. [PubMed]
13. Goodman E, Adler NE, Daniels SR, Morrison JA, Slap GB, Dolan LM. Impact of objective and subjective social status on obesity in a biracial cohort of adolescents. Obes Res. 2003;11:1018–1026. [PubMed]
14. Halder I, Muldoon MF, Ferrell RE, Manuck SB. Serotonin receptor 2A (HTR2A) gene polymorphisms are associated with blood pressure, central adiposity, and the metabolic syndrome. Metab Syndr Relat Disord. 2007;5:323–330. [PMC free article] [PubMed]
15. Bleil ME, Gianaros PJ, Jennings JR, Flory JD, Manuck SB. Trait negative affect: Toward an integrated model of understanding psychological risk for impairment in cardiac autonomic function. Psychosom Med. 2008;70:328–337. [PubMed]
16. Hall MH, Muldoon MF, Jennings JR, Buyssee DJ, Flory JD, Manuck SB. Self-reported sleep duration is associated with the metabolic syndrome in midlife adults. Sleep. 2008;31:635–643. [PubMed]
17. Muldoon MF, Nazzaro P, Sutton-Tyrrell K, Manuck SB. White-coat hypertension and carotid artery atherosclerosis: a matching study. Arch Intern Med. 2000;160:1507–1512. [PubMed]
18. Grundy SM. Metabolic syndrome scientific statement by the American Heart Association and the National Heart, Lung, and Blood Institute. Arterioscler Thromb Vasc Biol. 2005;25:2243–2244. [PubMed]
19. Grundy SM, Cleeman JL, Daniels SR, Donato KA, Eckel RH, Franklin BA. Diagnosis and management of the metabolic syndrome: An American Heart Association/National Heart Lung and Blood Institute Scientific Statement. Circulation. 2005;112:2735–2752. [PubMed]
20. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DR, Turner RC. Homeostasis model assessment: insulin resistance and β-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28:412–419. [PubMed]
21. Manuck SB, Bleil ME, Petersen KL, Flory JD, Mann JJ, Ferrell RE, Muldoon MF. The socio-economic status of communities predicts variation in brain serotonergic responsivity. Psychol Med. 2005;35:519–528. [PubMed]
22. Matthews KA, Flory JD, Muldoon MF, Manuck SB. Does socioeconomic status relate to central serotonergic responsivity in healthy adults? Psychosom Med. 2000;62:231–237. [PubMed]
23. Paffenbarger RS, Wing AL, Hyde RT. Physical activity as an index of heart attack risk in college alumni. Am J Epidemiol. 1978;108:161–175. [PubMed]
24. Paffenbarger RS, Hyde RT, Wing AL, Lee I, Jung DL, Kampert JB. The association of changes in physical-activity level and other lifestyle characteristics with mortality among men. N Engl J Med. 1993;328:538–545. [PubMed]
25. Ford ES, Giles WH, Dietz WH. Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey. JAMA. 2002;287:356–359. [PubMed]
26. Jeffery RW, Wing RR, Sherwood NE, Tate DF. Physical activity and weight loss: does prescribing higher physical activity goals improve outcome? Am J Clin Nutr. 2003;78:669–670. [PubMed]
27. Eckel RH, Grundy SM, Zimmet PZ. The metabolic syndrome. Lancet. 2005;365:1415–1428. [PubMed]
28. Gami A, Witt B, Howard D, Erwin P, Gami L, Somers V, Montori VM. Metabolic syndrome and risk of incident cardiovascular events and death: A systematic review and meta-analysis of longitudinal studies. J Am Coll Cardiol. 2007;49:403–414. [PubMed]
29. Cavaghan MK, Ehrmann DA, Polonsky KS. Interactions between insulin resistance and insulin secretion in the development of glucose intolerance. J Clin Invest. 2000;106:329–333. [PMC free article] [PubMed]
30. Haffner SB, Stern MP, Hazuda HP, Mitchell BD, Patterson JK. Cardiovascular risk factors in confirmed prediabetic individuals. Does the clock for coronary heart disease start ticking before the onset of clinical diabetes? JAMA. 1990;263:2893–2899. [PubMed]
31. Ghaed SG, Gallo LC. Subjective social status, objective socioeconomic status, and cardiovascular risk in women. Health Psychology. 2007;26:668–674. [PubMed]
32. Kopp M, Skrabski A, Rethelyi J, Kawachi I, Adler NE. Self-rated health, subjective social status and middle-aged mortality in a changing society. Behav Med. 2004;30:65–70. [PubMed]
33. de Castro AB, Gee GC, Takeuchi DT. Examining alternative measures of social disadvantage among Asian Americans: The relevance of economic opportunity, subjective social status, and financial strain for health. J Immigrant Minority Health. 2009 DOI 10.1007/s10903-009-9258-3. [PMC free article] [PubMed]
34. Ritterman ML, Fernald LC, Ozer EJ, Adler NE, Gutierrez JP, Syme SL. Objective and subjective social class gradients for substance use among Mexican adolescents. Soc Sci Med. 2009;68:1843–1851. [PubMed]
35. Gianaros PJ, Horenstein JA, Cohen S, Matthews KA, Brown SM, Flory JD, Critchley HD, Manuck SB, Hariri AR. Perigenual anterior cingulated morphology covaries with perceived social standing. Soc Cogn Affect Neurosci. 2007;2:161–173. [PMC free article] [PubMed]
36. Gallo LC, Matthews KA. Understanding the association between socioeconomic status and physical health: Do negative emotions play a role? Psychol Bull. 2003;129:10–51. [PubMed]
37. Adler NE, Snibbe AC. The role of psychosocial processes in explaining the gradient between socioeconomic status and health. Curr Dir Psychol Sci. 2003;12:119–123.
38. Wright CE, Steptoe A. Subjective socioeconomic position, gender and cortisol responses to waking in an elderly population. Psychoneuroendocrinology. 2005;30:582–590. [PubMed]
39. Drevets WC, Ongur D, Price JL. Neuroimaging abnormalities in the subgenual prefrontal cortex: implications for the pathophysiology of familial mood disorders. Mol Psychiatry. 1998;3:220–226. [PubMed]
40. Brunner EJ, Hemingway H, Walker BR, Page M, Clarke P, Juneja M, Shipley MJ, Kumari M, Andrew R, Seckl JR, Papadopoulos A, Checkley S, Rumley A, Lowe GDO, Stansfeld SA, Marmot MG. Adrenocortical, autonomic, and inflammatory causes of the metabolic syndrome. Circulation. 2002;106:2659–2665. [PubMed]
41. Pasquali R, Valentina V, Cacciari M, Pagotto U. The hypothalamic-pituitary, adrenal axis activity in obesity and the metabolic syndrome. Ann N Y Acad Sci. 2006;1083:111–128. [PubMed]
42. Walker BR, Phillips DIW, Noon JP, Panarelli M, Andrew R, Edwards HV, Holton DW, Seckl JR, Webb DJ, Watt GCM. Increased glucocorticoid activity in men with cardiovascular risk factors. Hypertension. 1998;31:891–895. [PubMed]
43. Walker BR, Soderberg S, Lindahl B, Olsson T. Independent effects of obesity and cortisol in predicting cardiovascular risk factors in men and women. J Intern Med. 2000;247:198–204. [PubMed]
44. Wilson PW, D’Agostino RB, Parise H, Sullivan L, Meigs JB. Metabolic syndrome as a precursor of cardiovascular disease and type 2 diabetes mellitus. Circulation. 2005;112:3066–3072. [PubMed]