Abstract
Aims/Introduction
Diabetes, hypertension, and obesity increase the risk of chronic kidney disease and associated mortality. Race and socioeconomic status (SES) differences in the effects of these risk factors are, however, still unknown. The current study aimed to investigate whether or not race and SES alter the effects of diabetes, hypertension, and obesity on mortality due to renal disease.
Materials and Methods
Data came from the Americans' Changing Lives Study, 1986–2011, a nationally representative prospective cohort of adults with 25 years of follow up. The study included 3,361 adults aged 25 years and older who were followed for up to 25 years. The outcome was death from renal disease. Diabetes, hypertension, and obesity were the main predictors. Race and SES (education, income, and employment) were moderators. Health behaviors and health status at baseline were covariates. We used Cox proportional hazards models for data analysis.
Results
In separate models, diabetes, hypertension, and obesity at baseline were associated with a higher risk of death from renal disease. From our SES indicators, education and income interacted with diabetes, hypertension, and obesity on death from renal disease. In a consistent pattern, diabetes, hypertension, and obesity showed stronger effects on the risk of death from renal disease among high‐SES groups compared with low‐SES individuals. Race and employment did not alter the effects of diabetes, hypertension and obesity on the risk of death from renal disease.
Conclusions
Social groups differ in how diabetes, hypertension, and obesity influence health outcomes over long‐term periods. Elimination of disparities in renal disease mortality in the USA requires understanding of the complex and non‐linear effects of socioeconomic and medical risk factors on health outcomes. Multidisciplinary programs and policies are required to reduce social inequality in renal disease burden caused by diabetes, hypertension, and obesity.
Keywords: Diabetes, Renal diseases, Social class
Introduction
Low socioeconomic status (SES) is a major risk factor for kidney disease1, 2. Minority status also increases the risk of death from renal disease3. In both sexes and all racial groups in the USA, the incidence of end‐stage renal disease (ESRD) by all primary renal diseases is greatest in the lowest SES score quartile, and decreases progressively as SES increases2. African Americans are three‐to‐four times as likely to develop kidney failure compared with White Americans4. Despite composing just 13% of the population, one‐third of all kidney failures in the USA occur among African Americans4.
In the USA, race and SES operate jointly, as minorities tend to have lower SES5, 6. For instance, race is closely associated with education, employment, and income6. However, a complex association exists between race, SES, risk factors (diabetes [DM], hypertension [HTN], and obesity), and renal disease mortality7, 8.
Medical risk factors, such as DM, obesity, and HTN, increase the risk of renal disease mortality9, 10, 11, 12, 13. Diabetes, HTN, and obesity might explain why race and SES predict renal disease mortality7, 8. Such medical risk factors9, 10 are well‐known causes of kidney disease, and at the same time are more common among low‐SES groups14, 15and African Americans12, 16, 17.
The medical risk factors of chronic kidney disease (CKD), such as DM12, obesity13, and HTN11, increase the risk of mortality as a result of renal diseases over long periods of time. The effect of DM on the development and burden of chronic kidney disease is well established18. In fact, chronic kidney disease is a known complication of DM19. Hypertension is also a known etiological risk factor for the development of chronic kidney disease20, as the risk of end‐stage renal disease increases in the presence of HTN20. Individuals with obesity also develop kidney disease, partly because of the higher risk of DM and HTN13.
To build on the existing literature on the complex effects of race, SES, DM, HTN, and obesity on renal disease outcomes7, 8, we examined race and SES (education, income, and employment) differences in the effects of three established medical risk factors (DM, HTN, and obesity) on mortality as a result of renal diseases over 25 years in the USA.
Methods
Setting
Data came from the Americans' Changing Lives (ACL) Study, a nationally‐representative USA cohort study carried out from 1986 until 2011. Detailed information on the study design is available elsewhere21, 22.
Sampling and participants
The ACL enrolled a stratified multistage probability sample of adults aged 25 years or older who lived in the continental USA in 1986. The study included 3,617 non‐institutionalized respondents (representing 70% of sampled households and 68% of sample individuals at baseline) with an oversampling of those age 60 years and older, and African Americans. Wave 1 included 70% of sampled households and 68% of sampled individuals. Further interviews were carried out in 1989, 1994, 2001/2 and 2011, but information from those interviews was not relevant for these analyses.
Informed consent
All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) with the Helsinki Declaration of 1975, as revised in 2000. Informed consent was obtained from all participants included in the study.
Measures
Data on demographics, SES, medical risk factors, health behaviors, and health status were measured at baseline during face‐to‐face interviews in 1986.
Moderators
Race
Race was defined as non‐Hispanic Black or non‐Hispanic White based on a coding of self‐reported items asking about Hispanic ethnicity, nativity, and racial category.
SES
Data on SES were measured with an indicator of education (<12 years of education, ≥12 years), income (10‐level categorical variable consisting of <$5,000, $5,000–9,000, $10,000–14,000, $15,000–19,000, $20,000–24,000, $25,000–29,000, $30,000–39,000, $40,000–59,000, $60,000–79,000, $80,000+) and employment (employed, unemployed) at baseline.
Covariates
Demographics
Demographic data included sex (a dichotomous variable with male as the referent category) and age (a continuous variable as number of years since birth).
Health behaviors
We collected data on self‐reported smoking (current smoker vs other), drinking (any drinking vs other), and exercise (any exercise vs other) using the following single‐item measures. ‘Do you smoke cigarettes now?’ ‘Do you ever drink beer, wine, or liquor?’ ‘How often do you engage in active sports or exercise – would you say often, sometimes, rarely, or never?’ Responses of the first two items were ‘yes’ and ‘no,’ and responses of the third item were: (i) often; (ii) sometimes; (iii) rarely; and (iv) never.
Self‐rated health
Self‐rated health (SRH) was a single‐item scale on subjective overall health, with five categories: poor, fair, good, very good, and excellent. SRH was treated as a dichotomous measure, where the original five categories were collapsed to the poor/fair vs good/very good/excellent categories. Categorizing SRH in this way has shown acceptable test–retest reliability and validity23.
Depressive symptoms
We used the 11‐item Center for Epidemiological Studies‐Depression scale24. In this shortened version of the Center for Epidemiological Studies‐Depression scale, a score of 1–3 is given to each item. We used the mean item score as the total score, ranging from 1 to 3. The abbreviated Center for Epidemiological Studies‐Depression scale has been used in other studies, and has shown acceptable reliability and validity25.
Predictors
DM
Self‐reported DM was assessed at baseline. All participants were asked whether a healthcare provider had ever told them they had DM22, 26.
HTN
Self‐reported HTN was assessed at baseline. All participants were asked whether a healthcare provider had ever told them they had HTN22, 26.
Obesity
The body mass index (BMI) was calculated based on self‐reported weights and heights. Weight and height were originally collected in pounds (1 pound = 0.453 kg) and feet (1 foot = 0.3048 m) / inches (1 inch = 0.0254 m), respectively. Obesity was defined as BMI ≥30 kg/m2 27. BMI calculated based on self‐reported weight and height is known to be closely correlated with BMI based on direct measures of height and weight28. However, using self‐reported weight and height could lead to some degrees of underestimation of BMI29, due to a systematic tendency for humans to underestimate their weight and to overestimate their height30.
Outcome
Mortality as a result of renal diseases
The main outcome variables were mortality from renal diseases and time of death. Information on all deaths from mid‐1986 through 2011 was obtained through the National Death Index, death certificates and also from informants. In most cases, time and cause of death were verified with death certificates. The handful of cases where death could not be verified with death certificates was reviewed carefully, and actual death was certain in all cases. Only in these cases was the date of death ascertained from the informants or the National Death Index report, rather than the death certificate26, 31. Cause of death was coded as unknown if the death certificate or National Death Index report were unavailable.
We used the International Classification of Diseases (ICD)‐9 and ICD‐10 codes32, 33, whichever was current at the time the death was recorded, to determine death from renal diseases (kidney–urinary). For ICD‐9 codes, we used codes 650 (acute glomerulonephritis and nephrotic syndrome), 660 (chronic glomerulonephritis, nephritis and nephropathy, not specified as acute or chronic, and renal sclerosis, unspecified), 670 (renal failure, disorders resulting from impaired renal function and small kidney of unknown causes), 680 (infections of kidney) and 690 (hyperplasia of prostate). For ICD‐10 codes, we used the categorization of 113 selected causes of death provided by the World Health Organization, for which codes 97 (nephritis, nephrotic syndrome and nephrosis), 98 (acute and rapidly progressive nephritic and nephrotic syndrome), 99 (chronic glomerulonephritis, nephritis and nephropathy not specified as acute or chronic, and renal sclerosis unspecified), 100 (renal failure), 101 (other disorders of kidney), 102 (infections of kidney), 103 (hyperplasia of prostate) and 104 (inflammatory diseases of female pelvic organ) were used. Respondents who died of other causes were censored at the time of death. Time of death was registered as the number of months from time of enrollment in the study to time of death, based on the month of death and the month of the baseline interview.
Statistical analysis
Univariate, bivariate, and multivariable analyses were carried out using Stata 13.0 (StataCorp, College Station, Texas, USA). We used Taylor series linearization to estimate standard errors as a result of stratification and clustering, as well as non‐response. Hazard ratios with 95% confidence intervals (CI) are reported. For main effects, P < 0.05 was considered statistically significant. For interactions, we considered P < 0.1 as statistically significant.
Cox proportional hazards models require two variables, a binary outcome (renal death) and also time to the renal death event, or to censoring. Time to event was defined as the number of months from baseline to death, loss to follow up, or the end of the year 2011. Renal death was coded as zero if the respondent did not die, or died from any other causes.
For multivariable analysis, 24 Cox proportional hazards models were fitted to the data (six for each moderator). At each step, we ran models with main effects (model 1a, model 2a, and model 3a). Then we ran models with the interaction terms (model 1b, model 2b, and model 3b). The outcome was death from renal diseases and the time that outcome occurred since baseline (time to death from renal diseases). Race and SES indicators (education, income, employment) were the moderators. Chronic medical diseases (DM, HTN, obesity) were predictors. Health behaviors (smoking, drinking, and exercise) and health status (SRH and depressive symptoms) at baseline were potential confounders.
Results
Table 1 shows descriptive statistics for the overall sample, and based on race. Tables 2, 3, 4, 5 each show six Cox proportional hazard regression models, with models on the left being the main effect models, and models on the right testing the interactions.
Table 1.
Descriptive statistics of demographic, socioeconomic, medical conditions, and health behaviors overall and among Blacks and Whites at baseline
| All | Whites | Blacks | ||||
|---|---|---|---|---|---|---|
| Mean (SE) | 95% CI | Mean (SE) | 95% CI | Mean (SE) | 95% CI | |
| Age | 47.79 (0.53) | 46.72–48.86 | 47.98 (0.60) | 46.77–49.19 | 46.37 (0.71) | 44.93–47.81 |
| Income* | 5.41 (0.09) | 5.22–5.60 | 5.57 (0.10) | 5.36–5.77 | 4.25 (0.18) | 3.88–4.62 |
| Exercise* | 0.02 (0.03) | –0.03–0.07 | 0.06 (0.03) | 0.00–0.11 | –0.22 (0.05) | –0.33–0.12 |
| % (SE) | 95% CI | % (SE) | 95% CI | % (SE) | 95% CI | |
| Sex | ||||||
| Male | 47.26 (0.01) | 44.86–49.68 | 47.82 (0.01) | 45.12–50.52 | 43.18 (0.02) | 38.79–47.69 |
| Female | 52.74 (0.01) | 50.32–55.14 | 52.18 (0.01) | 49.48–54.88 | 56.82 (0.02) | 52.31–61.21 |
| Education* | 12.53 (0.10) | 12.34–12.73 | 12.69 (0.11) | 12.48–12.90 | 11.37 (0.23) | 10.90–11.84 |
| % (SE) | 95% CI | % (SE) | 95% CI | % (SE) | 95% CI | |
| HTN* | 21.37 (0.01) | 19.67–23.17 | 19.77 (0.01) | 18.11–21.53 | 33.17 (0.03) | 28.20–38.54 |
| DM* | 5.73 (0.00) | 4.80–6.82 | 5.25 (0.01) | 4.24–6.50 | 9.22 (0.01) | 7.75–10.95 |
| Obesity* | 14.47 (0.01) | 12.86–16.24 | 13.52 (0.01) | 11.72–15.54 | 21.45 (0.02) | 17.88–25.52 |
| Smoking* | 30.45 (0.01) | 27.81–33.23 | 29.70 (0.01) | 26.85–32.72 | 35.98 (0.03) | 30.81–41.49 |
| Drinking* | 60.02 (0.02) | 56.68–63.26 | 61.50 (0.02) | 58.06–64.83 | 49.10 (0.03) | 43.55–54.68 |
| Renal death | 0.52 (0.00) | 0.26–1.03 | 0.44 (0.00) | 0.18–1.06 | 1.15 (0.00) | 0.63–02.09 |
*P < 0.05. CES‐D, Center for Epidemiologic Studies Depression Scale; CI, confidence interval; CMC, chronic medical conditions; DM, diabetes mellitus; HTN, hypertension; SE, standard error.
Table 2.
Summary of Cox regressions on the interaction between low education and medical risk factors of death from renal disease in the USA
| HR (SE) | 95% CI | P | HR (SE) | 95% CI | P | |
|---|---|---|---|---|---|---|
| Model a (main effects) | Model b (interactions) | |||||
| Model 1 | ||||||
| Race (Black) | 2.20 (0.96) | 0.92–5.29 | 0.076 | 2.13 (0.92) | 0.89–5.08 | 0.088 |
| Age (years) | 1.09 (0.02) | 1.06–1.13 | 0.000 | 1.09 (0.02) | 1.06–1.12 | 0.000 |
| Sex (women) | 0.40 (0.25) | 0.12–1.39 | 0.146 | 0.51 (0.28) | 0.17–1.51 | 0.220 |
| SES (low education) | 0.56 (0.30) | 0.19–1.66 | 0.292 | 1.45 (0.93) | 0.40–5.26 | 0.566 |
| Smoking | 3.46 (1.86) | 1.17–10.24 | 0.026 | 3.32 (1.61) | 1.25–8.81 | 0.017 |
| Drinking | 0.13 (0.08) | 0.04–0.42 | 0.001 | 0.16 (0.08) | 0.06–0.44 | 0.001 |
| Exercise | 0.78 (0.17) | 0.51–1.21 | 0.269 | 0.85 (0.17) | 0.57–1.26 | 0.414 |
| SRH (poor) | 2.26 (1.03) | 0.90–5.68 | 0.082 | 2.20 (0.99) | 0.89–5.44 | 0.085 |
| Depressive symptoms | 1.52 (0.27) | 1.06–2.17 | 0.024 | 1.60 (0.28) | 1.12–2.29 | 0.011 |
| DM | 7.57 (4.02) | 2.60–22.07 | 0.000 | 18.90 (14.83) | 3.89–91.81 | 0.001 |
| DM × SES | – | – | – | 0.12 (0.12) | 0.02–0.92 | 0.042 |
| Model 2 | ||||||
| Race (Black) | 1.84 (0.87) | 0.71–4.76 | 0.202 | 1.99 (0.92) | 0.78–5.07 | 0.144 |
| Age (years) | 1.08 (0.01) | 1.05–1.11 | 0.000 | 1.08 (0.01) | 1.05–1.11 | 0.000 |
| Sex (women) | 0.36 (0.25) | 0.09–1.49 | 0.155 | 0.41 (0.28) | 0.11–1.60 | 0.195 |
| SES (low education) | 0.55 (0.35) | 0.15–1.95 | 0.346 | 7.89 (7.49) | 1.17–53.37 | 0.035 |
| Smoking | 3.73 (2.13) | 1.18–11.78 | 0.026 | 3.68 (2.05) | 1.20–11.32 | 0.024 |
| Drinking | 0.12 (0.08) | 0.03–0.47 | 0.003 | 0.13 (0.08) | 0.03–0.47 | 0.003 |
| Exercise | 0.81 (0.18) | 0.52–1.26 | 0.341 | 0.87 (0.19) | 0.57–1.34 | 0.524 |
| SRH (poor) | 2.40 (1.14) | 0.92–6.27 | 0.073 | 2.42 (1.08) | 0.98–5.95 | 0.055 |
| Depressive symptoms | 1.54 (0.22) | 1.15–2.06 | 0.005 | 1.58 (0.25) | 1.15–2.17 | 0.006 |
| HTN | 4.25 (2.69) | 1.19–15.18 | 0.027 | 29.42 (28.03) | 4.32–200.48 | 0.001 |
| HTN × SES | – | – | – | 0.03 (0.03) | 0.00–0.24 | 0.002 |
| Model 3 | ||||||
| Race (Black) | 1.93 (1.00) | 0.68–5.50 | 0.213 | 1.97 (0.96) | 0.74–5.24 | 0.168 |
| Age (year) | 1.10 (0.01) | 1.07–1.13 | 0.000 | 1.10 (0.02) | 1.07–1.13 | 0.000 |
| Sex (women) | 0.36 (0.25) | 0.09–1.42 | 0.141 | 0.42 (0.28) | 0.11–1.62 | 0.204 |
| SES (low education) | 0.59 (0.35) | 0.17–1.97 | 0.378 | 1.48 (0.91) | 0.43–5.11 | 0.524 |
| Smoking | 3.41 (1.96) | 1.07–10.86 | 0.039 | 3.57 (2.04) | 1.12–11.32 | 0.032 |
| Drinking | 0.13 (0.08) | 0.04–0.42 | 0.001 | 0.15 (0.08) | 0.05–0.41 | 0.001 |
| Exercise | 0.82 (0.16) | 0.55–1.23 | 0.332 | 0.87 (0.17) | 0.58–1.29 | 0.473 |
| SRH (poor) | 2.63 (1.16) | 1.08–6.38 | 0.033 | 2.38 (0.92) | 1.09–5.17 | 0.030 |
| Depressive symptoms | 1.51 (0.19) | 1.17–1.95 | 0.002 | 1.59 (0.24) | 1.17–2.15 | 0.004 |
| Obesity | 2.88 (1.71) | 0.87–9.51 | 0.082 | 7.14 (5.42) | 1.54–32.97 | 0.013 |
| Obesity × SES | – | – | – | 0.11 (0.11) | 0.02–0.79 | 0.029 |
CI, confidence interval; DM; diabetes, HR, hazard ratio; HTN; hypertension; SE, standard error; SES, socioeconomic status; SRH; self‐rated health.
Table 3.
Summary of Cox regressions on the interaction between unemployment and medical risk factors of death from renal disease in the USA
| HR (SE) | 95% CI | P | HR (SE) | 95% CI | P | |
|---|---|---|---|---|---|---|
| Model a (main effects) | Model b (interactions) | |||||
| Model 1 | ||||||
| Race (Black) | 1.99 (0.95) | 0.76–5.21 | 0.156 | 1.99 (0.96) | 0.75–5.27 | 0.160 |
| Age (years) | 1.08 (0.02) | 1.03–1.12 | 0.001 | 1.08 (0.02) | 1.03–1.13 | 0.001 |
| Sex (women) | 0.35 (0.25) | 0.08–1.51 | 0.154 | 0.35 (0.25) | 0.08–1.50 | 0.152 |
| SES (unemployed) | 2.29 (1.35) | 0.70–7.50 | 0.166 | 2.43 (1.93) | 0.50–11.98 | 0.266 |
| Smoking | 3.26 (1.72) | 1.13–9.42 | 0.030 | 3.24 (1.79) | 1.07–9.83 | 0.039 |
| Drinking | 0.15 (0.08) | 0.05–0.44 | 0.001 | 0.15 (0.08) | 0.05–0.43 | 0.001 |
| Exercise | 0.76 (0.17) | 0.48–1.19 | 0.220 | 0.76 (0.17) | 0.49–1.18 | 0.217 |
| SRH (poor) | 1.80 (0.71) | 0.81–4.00 | 0.142 | 1.81 (0.71) | 0.82–3.98 | 0.137 |
| Depressive Symptoms | 1.42 (0.26) | 0.98–2.06 | 0.065 | 1.42 (0.27) | 0.97–2.07 | 0.072 |
| DM | 6.83 (3.55) | 2.39–19.45 | 0.001 | 7.98 (9.84) | 0.67–95.68 | 0.099 |
| DM × SES | – | – | – | 0.83 (1.20) | 0.05–15.10 | 0.900 |
| Model 2 | ||||||
| Race (Black) | 1.70 (0.93) | 0.57–5.12 | 0.335 | 1.67 (0.92) | 0.55–5.05 | 0.359 |
| Age (years) | 1.06 (0.02) | 1.02–1.11 | 0.009 | 1.06 (0.02) | 1.02–1.11 | 0.008 |
| Sex (women) | 0.32 (0.25) | 0.06–1.57 | 0.155 | 0.31 (0.25) | 0.06–1.55 | 0.150 |
| SES (unemployed) | 2.71 (1.74) | 0.74–9.90 | 0.127 | 5.22 (4.98) | 0.76–35.65 | 0.090 |
| Smoking | 3.39 (1.78) | 1.18–9.74 | 0.024 | 3.34 (1.78) | 1.14–9.78 | 0.028 |
| Drinking | 0.15 (0.08) | 0.05–0.45 | 0.001 | 0.15 (0.08) | 0.05–0.46 | 0.001 |
| Exercise | 0.78 (0.18) | 0.49–1.24 | 0.288 | 0.78 (0.18) | 0.49–1.24 | 0.284 |
| SRH (poor) | 1.89 (0.76) | 0.84–4.25 | 0.118 | 1.92 (0.75) | 0.87–4.24 | 0.103 |
| Depressive symptoms | 1.46 (0.22) | 1.08–1.96 | 0.015 | 1.46 (0.21) | 1.09–1.94 | 0.011 |
| HTN | 3.82 (2.21) | 1.19–12.24 | 0.025 | 8.67 (8.90) | 1.10–68.59 | 0.041 |
| HTN × SES | – | – | – | 0.36 (0.46) | 0.03–4.89 | 0.433 |
| Model 3 | ||||||
| Race (Black) | 1.81 (1.09) | 0.54–6.08 | 0.330 | 1.84 (1.10) | 0.55–6.14 | 0.314 |
| Age (years) | 1.08 (0.02) | 1.04–1.12 | 0.000 | 1.08 (0.02) | 1.04–1.12 | 0.000 |
| Sex (Women) | 0.31 (0.25) | 0.06–1.54 | 0.147 | 0.31 (0.24) | 0.06–1.52 | 0.144 |
| SES (unemployed) | 2.99 (2.14) | 0.71–12.62 | 0.133 | 2.19 (1.25) | 0.69–6.91 | 0.176 |
| Smoking | 3.16 (1.72) | 1.06–9.43 | 0.039 | 3.19 (1.73) | 1.07–9.52 | 0.038 |
| Drinking | 0.16 (0.08) | 0.06–0.43 | 0.001 | 0.15 (0.08) | 0.06–0.43 | 0.001 |
| Exercise | 0.80 (0.17) | 0.52–1.21 | 0.278 | 0.79 (0.16) | 0.53–1.20 | 0.270 |
| SRH (poor) | 2.05 (0.74) | 0.99–4.25 | 0.052 | 2.04 (0.75) | 0.97–4.28 | 0.059 |
| Depressive symptoms | 1.45 (0.21) | 1.09–1.93 | 0.011 | 1.45 (0.21) | 1.08–1.95 | 0.014 |
| Obesity | 2.75 (1.64) | 0.83–9.13 | 0.097 | 1.29 (0.98) | 0.28–5.98 | 0.736 |
| Obesity × SES | – | – | – | 2.41 (2.46) | 0.31–18.87 | 0.394 |
CI, confidence interval; DM; diabetes, HR, hazard ratio; HTN; hypertension; SE, standard error; SES, socioeconomic status; SRH; self‐rated health.
Table 4.
Summary of Cox regressions on the interaction between low income and medical risk factors of death from renal disease in the USA
| HR (SE) | 95% CI | P | HR (SE) | 95% CI | P | |
|---|---|---|---|---|---|---|
| Model a (main effects) | Model b (interactions) | |||||
| Model 1 | ||||||
| Race (Black) | 2.03 (0.86) | 0.87–4.75 | 0.099 | 2.04 (0.86) | 0.87–4.74 | 0.097 |
| Age (years) | 1.09 (0.02) | 1.06–1.13 | 0.000 | 1.09 (0.02) | 1.05–1.12 | 0.000 |
| Sex (women) | 0.40 (0.26) | 0.11–1.50 | 0.169 | 0.45 (0.27) | 0.14–1.49 | 0.186 |
| SES (Low Income) | 0.80 (0.49) | 0.23–2.74 | 0.717 | 1.84 (1.45) | 0.38–9.03 | 0.442 |
| Smoking | 3.42 (1.90) | 1.11–10.50 | 0.033 | 3.46 (1.78) | 1.23–9.75 | 0.020 |
| Drinking | 0.13 (0.08) | 0.04–0.44 | 0.002 | 0.15 (0.07) | 0.06–0.41 | 0.000 |
| Exercise | 0.78 (0.17) | 0.50–1.19 | 0.239 | 0.81 (0.16) | 0.54–1.20 | 0.283 |
| SRH (poor) | 2.14 (1.00) | 0.83–5.48 | 0.112 | 1.96 (0.85) | 0.82–4.69 | 0.125 |
| Depressive symptoms | 1.46 (0.25) | 1.04–2.06 | 0.031 | 1.46 (0.26) | 1.02–2.08 | 0.040 |
| DM | 7.35 (3.88) | 2.53–21.30 | 0.000 | 19.57 (17.64) | 3.18–120.25 | 0.002 |
| DM × SES | – | – | – | 0.20 (0.19) | 0.03–1.29 | 0.090 |
| Model 2 | ||||||
| Race (Black) | 1.75 (0.80) | 0.70–4.41 | 0.227 | 1.80 (0.82) | 0.72–4.51 | 0.202 |
| Age (years) | 1.08 (0.02) | 1.05–1.11 | 0.000 | 1.08 (0.01) | 1.05–1.11 | 0.000 |
| Sex (women) | 0.39 (0.25) | 0.11–1.44 | 0.154 | 0.42 (0.26) | 0.13–1.43 | 0.162 |
| SES (low income) | 0.60 (0.46) | 0.13–2.77 | 0.508 | 3.56 (3.19) | 0.59–21.58 | 0.163 |
| Smoking | 3.66 (2.06) | 1.18–11.38 | 0.026 | 3.77 (2.09) | 1.24–11.49 | 0.021 |
| Drinking | 0.13 (0.08) | 0.03–0.48 | 0.003 | 0.13 (0.08) | 0.03–0.48 | 0.003 |
| Exercise | 0.80 (0.17) | 0.53–1.23 | 0.304 | 0.80 (0.17) | 0.52–1.23 | 0.295 |
| SRH (poor) | 2.33 (1.11) | 0.89–6.09 | 0.083 | 2.29 (1.02) | 0.94–5.60 | 0.068 |
| Depressive Symptoms | 1.54 (0.23) | 1.15–2.07 | 0.005 | 1.60 (0.26) | 1.16–2.20 | 0.005 |
| HTN | 4.28 (2.83) | 1.13–16.23 | 0.033 | 19.30 (18.48) | 2.81–132.72 | 0.003 |
| HTN × SES | – | – | – | 0.09 (0.10) | 0.01–0.81 | 0.033 |
| Model 3 | ||||||
| Race (Black) | 1.82 (0.91) | 0.67–4.98 | 0.235 | 2.03 (0.90) | 0.83–4.95 | 0.117 |
| Age (years) | 1.10 (0.02) | 1.07–1.13 | 0.000 | 1.10 (0.02) | 1.07–1.13 | 0.000 |
| Sex (women) | 0.38 (0.25) | 0.10–1.40 | 0.143 | 0.46 (0.25) | 0.15–1.40 | 0.166 |
| SES (low income) | 0.76 (0.48) | 0.21–2.72 | 0.662 | 2.40 (2.22) | 0.37–15.49 | 0.350 |
| Smoking | 3.34 (1.89) | 1.07–10.41 | 0.038 | 3.75 (2.17) | 1.17–12.01 | 0.027 |
| Drinking | 0.14 (0.08) | 0.04–0.42 | 0.001 | 0.16 (0.07) | 0.06–0.41 | 0.000 |
| Exercise | 0.82 (0.16) | 0.55–1.22 | 0.322 | 0.85 (0.16) | 0.59–1.24 | 0.391 |
| SRH (poor) | 2.56 (1.17) | 1.02–6.44 | 0.046 | 2.26 (0.85) | 1.06–4.82 | 0.035 |
| Depressive symptoms | 1.49 (0.18) | 1.17–1.91 | 0.002 | 1.50 (0.19) | 1.17–1.93 | 0.002 |
| Obesity | 2.80 (1.60) | 0.89–8.82 | 0.078 | 11.33 (11.68) | 1.42–90.38 | 0.023 |
| Obesity × SES | – | – | – | 0.10 (0.11) | 0.01–0.91 | 0.041 |
CI, confidence interval; DM; diabetes, HR, hazard ratio; HTN; hypertension; SE, standard error; SES, socioeconomic status; SRH; self‐rated health.
Table 5.
Summary of Cox regressions on the interaction between race and medical risk factors of death from renal disease in the USA
| HR (SE) | 95% CI | P | HR (SE) | 95% CI | P | |
|---|---|---|---|---|---|---|
| Model a (main effects) | Model b (interactions) | |||||
| Model 1 | ||||||
| Race (Black) | 1.95 (0.93) | 0.74–5.10 | 0.169 | 2.41 (1.16) | 0.91–6.36 | 0.075 |
| Age (years) | 1.09 (0.02) | 1.06–1.13 | 0.000 | 1.09 (0.02) | 1.06–1.13 | 0.000 |
| Sex (women) | 0.37 (0.27) | 0.09–1.61 | 0.182 | 0.38 (0.27) | 0.09–1.58 | 0.178 |
| Smoking | 3.38 (1.85) | 1.12–10.20 | 0.031 | 3.35 (1.80) | 1.14–9.86 | 0.029 |
| Drinking | 0.13 (0.08) | 0.04–0.45 | 0.002 | 0.14 (0.08) | 0.04–0.44 | 0.001 |
| Exercise | 0.77 (0.17) | 0.50–1.19 | 0.232 | 0.77 (0.16) | 0.50–1.19 | 0.229 |
| SRH (poor) | 2.09 (0.91) | 0.87–5.03 | 0.097 | 2.05 (0.87) | 0.87–4.81 | 0.097 |
| Depressive symptoms | 1.44 (0.26) | 1.00–2.07 | 0.048 | 1.43 (0.26) | 0.99–2.06 | 0.054 |
| DM | 7.46 (4.07) | 2.48–22.40 | 0.001 | 8.51 (5.64) | 2.24–32.30 | 0.002 |
| DM × race | – | – | – | 0.62 (0.52) | 0.12–3.38 | 0.576 |
| Model 2 | ||||||
| Race (Black) | 1.61 (0.92) | 0.51–5.06 | 0.409 | 1.89 (1.22) | 0.51–6.93 | 0.330 |
| Age (years) | 1.08 (0.02) | 1.04–1.11 | 0.000 | 1.08 (0.02) | 1.04–1.11 | 0.000 |
| Sex (women) | 0.35 (0.27) | 0.07–1.69 | 0.186 | 0.35 (0.27) | 0.07–1.69 | 0.186 |
| Smoking | 3.57 (1.93) | 1.20–10.60 | 0.023 | 3.55 (1.86) | 1.24–10.19 | 0.020 |
| Drinking | 0.13 (0.08) | 0.04–0.48 | 0.003 | 0.13 (0.08) | 0.04–0.47 | 0.003 |
| Exercise | 0.80 (0.17) | 0.52–1.24 | 0.310 | 0.80 (0.17) | 0.52–1.24 | 0.311 |
| SRH (poor) | 2.22 (0.96) | 0.92–5.32 | 0.074 | 2.22 (0.98) | 0.92–5.40 | 0.076 |
| Depressive symptoms | 1.49 (0.20) | 1.14–1.96 | 0.005 | 1.50 (0.20) | 1.13–1.97 | 0.005 |
| HTN | 4.14 (2.65) | 1.14–15.00 | 0.032 | 4.32 (3.22) | 0.97–19.36 | 0.055 |
| HTN × race | – | – | – | 0.81 (0.77) | 0.12–5.50 | 0.828 |
| Model 3 | ||||||
| Race (Black) | 1.71 (1.05) | 0.49–5.89 | 0.389 | 2.38 (1.28) | 0.80–7.05 | 0.115 |
| Age (years) | 1.09 (0.02) | 1.06–1.13 | 0.000 | 1.10 (0.02) | 1.07–1.13 | 0.000 |
| Sex (women) | 0.35 (0.26) | 0.08–1.57 | 0.167 | 0.36 (0.26) | 0.08–1.57 | 0.168 |
| Smoking | 3.32 (1.85) | 1.08–10.22 | 0.037 | 3.44 (2.11) | 1.00–11.84 | 0.050 |
| Drinking | 0.14 (0.08) | 0.04–0.43 | 0.001 | 0.14 (0.08) | 0.05–0.43 | 0.001 |
| Exercise | 0.82 (0.16) | 0.55–1.22 | 0.321 | 0.82 (0.16) | 0.55–1.23 | 0.325 |
| SRH (poor) | 2.48 (1.03) | 1.07–5.74 | 0.035 | 2.45 (1.01) | 1.06–5.63 | 0.036 |
| Depressive symptoms | 1.47 (0.19) | 1.13–1.92 | 0.005 | 1.46 (0.21) | 1.09–1.95 | 0.012 |
| Obesity | 2.86 (1.77) | 0.82–9.96 | 0.096 | 3.48 (2.84) | 0.67–17.98 | 0.133 |
| Obesity × race | – | – | – | 0.49 (0.57) | 0.05–5.08 | 0.545 |
CI, confidence interval; DM; diabetes, HR, hazard ratio; HTN; hypertension; SE, standard error; SES, socioeconomic status; SRH; self‐rated health.
Table 2 shows the results of six Cox proportional hazard regression models with low education as the SES indicator (moderator). In model 1a, DM was associated with a higher risk of death from renal disease, net of health behaviors and other health factors. Model 1b showed an interaction between education and DM, suggesting that the predictive role of DM on risk of death from renal disease is larger for those with high education, compared with those with low education. According to the results of model 2a, HTN was associated with a higher risk of death from renal disease, net of health behaviors and other health factors. Model 2b showed an interaction between education and HTN, suggesting that the predictive role of HTN on risk of death from renal disease is larger for individuals with high education, compared with those with low education. According to the results of model 3a, obesity was associated with a higher risk of death from renal disease, net of health behaviors and other health factors. Model 3b showed an interaction between education and obesity, suggesting that the predictive role of obesity on risk of death from renal disease depends on education, with a stronger predictive role for people with high education compared with those with low education (Table 2).
Table 3 shows a summary of the results of six Cox proportional hazard regression models with employment as the SES indicator (moderator). According to the results of model 1a, DM was associated with a higher risk of death from renal disease, net of health behaviors and other health factors. Model 1b did not show an interaction between employment and DM, suggesting that the effect of DM on risk of death from renal disease does not depend on employment status. According to the results of model 2a, HTN was associated with a higher risk of death from renal disease, net of health behaviors and other health factors. Model 2b did not show an interaction between employment and DM, suggesting that the effect of DM on risk of death from renal disease does not depend on employment status. According to the results of model 3a, obesity was associated with a higher risk of death from renal disease, net of health behaviors and other health factors. Model 3b did not show an interaction between employment and obesity, suggesting that the effect of obesity on risk of death from renal disease does not depend on employment status (Table 3).
Table 4 shows a summary of the results of six Cox proportional hazard regression models with income as the SES indicator (moderator). According to the results of model 1a, DM was associated with a higher risk of death from renal disease, net of health behaviors and other health factors. Model 1b showed an interaction between income and DM, suggesting that the predictive role of DM on risk of death from renal disease depends on income, with a better prediction role for those with high income, compared with those with low income. According to the results of model 2a, HTN was associated with a higher risk of death from renal disease, net of health behaviors and other health factors. Model 2b showed an interaction between low income and HTN, suggesting that the effect of HTN on the risk of death from renal disease depends on income, with HTN better predicting renal disease mortality for those with a high income, compared with their low‐income counterparts. According to the results of model 3a, obesity was associated with a higher risk of death from renal disease, net of health behaviors and other health factors. Model 3b showed an interaction between income and obesity, suggesting that the effect of obesity on risk of death from renal disease depends on income, with a better predictive role of obesity for high‐ vs low‐income individuals (Table 4).
Table 5 shows a summary of the results of six Cox proportional hazard regression models with race as the moderator. According to the results of model 1a, DM was associated with a higher risk of death from renal disease, net of health behaviors and other health factors. Model 1b did not show an interaction between race and DM, suggesting that the effect of DM on risk of death from renal disease does not depend on race. Similarly, although HTN and obesity were associated with the outcome (model 2a and model 3a), these associations did not depend on race (model 2b and model 3b; Table 5).
Discussion
Using 25‐year data, the current study showed that DM, HTN, and obesity interact with SES on renal disease mortality. We found that the predictive role of DM, HTN, and obesity are higher for individuals with high education and high income, compared with their low‐education and low‐income counterparts. This finding expands the current literature on how DM, HTN, obesity, and SES predict renal disease mortality7, 8. In a recent study by Assari7, SES (income) and medical disease (DM and HTN), as well as health behaviors (smoking, drinking, and exercise) explained the effect of race on death from renal disease. In another study by Assari8, DM, HTN, and other chronic medical conditions explained the effects of race and SES on all‐cause mortality. None of these studies tested the differential effects of DM, HTN, and obesity on mortality based on race or SES, which is the unique contribution of the present study.
Diabetes and HTN are leading causes of kidney failure in the USA, particularly among minorities and low SES individuals3. Thus, the presence of risk factors, such as HTN11, DM12, obesity13, and other chronic medical conditions9, 10 convey information on future risk of mortality from renal disease in the USA. It was previously known that these medical risk factors are more common among racial minorities and low‐SES individuals12, 16, 17. The current study shows that role of these conditions as a predictor of mortality from renal disease is not universal across all subpopulations.
How DM increases the risk of chronic kidney disease is well‐established18, as studies have shown that chronic kidney disease is a complication of DM19. The association between chronic kidney disease and HTN is also bidirectional, as HTN is listed as a known etiological factor for the development of chronic kidney disease20. Longitudinal studies have consistently shown that baseline HTN causally contributes to the subsequent risk of end‐stage renal disease12. Based on the current study, DM and HTN better predict renal disease mortality for high SES people.
Several studies have tried to deconstruct the effects of race, SES, and medical risk factors on health34, 35, 36. These associations are, however, shown to be multiplicative rather than additive37, 38. There is still a debate on whether HTN, DM, and other conditions explain the effects of race and SES on health or not35, 36. Another remaining question is whether race and SES have parallel or overlapping effects39, 40, and whether their effects are simply because of increasing exposure to different risk factors41, 42 or also altering vulnerabilities to the effects of risk factors43. Race and SES interact with health outcomes44, 45, 46, as racial differences in health depend on SES44, and SES effects on health depend on race45, 46. Thus, race, SES, and other risk factors might have multiplicative rather than additive effects on health44. Findings of the current study support such an argument.
According to our findings, SES differences in death from renal disease are not exclusively a result of SES gaps in exposures to medical risk factors (DM, HTN, and obesity). Instead, the same risk factors can have different roles across social groups, which is in line with the differential effect hypothesis47, 48. Thus, understanding the causal paths between social factors, medical risk factors, and health outcomes requires understanding the existing complex heterogeneity in the effects of the same risk factors on outcomes43. We argue that such heterogeneity might be a result of contextual factors, such as race, ethnicity, sex, and place, that shape vulnerabilities and resilience to the effects of the same risk factors on outcomes49, 50, 51, 52.
Thus, it is not SES or medical risk factors, such as DM, HTN, and obesity, per se that cause renal disease20. As the distribution of risk factors depends on racial and SES groups16, differential exposure might ultimately result in differential vulnerability to the effects of risk factors43. Thus, disparities in kidney disease mortality in the USA are the result of complex interactions between social and medical risk factors, such as DM and HTN12, 53, 54, 55.
The present study had a few limitations that should be considered before the results are interpreted. The first limitation was a lack of measurement of kidney disease at baseline or over the course of the follow up. The second limitation was the measurement of HTN, DM, and obesity using self‐reported data, which is subject to measurement bias56. Self‐rated medical conditions should always be validated by other sources of information, such as medical records. The third limitation was that the present study did not measure decline in kidney function over time. The fourth limitation was the lack of information on the diabetes type. Type 1 and type 2 diabetes have different pathogenesis mechanisms, and the patterns of renal disease development and progression might also be different. Fifth, we did not have enough data to explore the role of other races and sex in the modification of the effect of medical risk factors on renal disease mortality in the present study. Despite these limitations, the current study made a unique contribution to the literature. The major strengths of the present study were long‐term follow up, large sample size, and a nationally representative USA sample.
To conclude, we found that SES (education and income) alters how medical risk factors (DM, HTN, and obesity) increase the risk of deaths from renal diseases over a 25‐year period, above and beyond health behaviors (smoking, drinking, and exercise) and other health factors (SRH and depressive symptoms). Thus, the effects of social and medical determinants of health in shaping the burden of chronic kidney disease are not additive, but multiplicative. These findings extend the existing knowledge on the complex link between race, SES, chronic disease, and renal disease mortality in the USA. Thus, the attribution of racial and SES differences in renal disease mortality to racial and SES differences in exposure to risk factors might be oversimplistic57.
Disclosure
The authors declare no conflict of interest.
Acknowledgments
Shervin Assari is supported by the Heinz C. Prechter Bipolar Research as well as Richard Tam Foundation at the University of Michigan Depression Center. The ACL study was supported by Grant #AG018418 from the National Institute on Aging (DHHS/NIH), and per the NIH Public Access Policy requires that peer‐reviewed research publications generated with NIH support are made available to the public through PubMed Central. NIH is not responsible for the data collection or analyses represented in this article. The ACL study was carried out by the Institute of Social Research, University of Michigan.
J Diabetes Investig 2017; 8: 590–599
References
- 1. Perneger TV, Whelton PK, Klag MJ. Race and end‐stage renal disease. Socioeconomic status and access to health care as mediating factors. Arch Intern Med 1995; 155: 1201–1208. [PubMed] [Google Scholar]
- 2. Ward MM. Socioeconomic status and the incidence of ESRD. Am J Kidney Dis 2008; 51: 563–572. [DOI] [PubMed] [Google Scholar]
- 3. U.S. Renal Data System , USRDS 2010 Annual Data Report: Atlas of Chronic Kidney Disease and End‐Stage Renal Disease in the United States, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, 2010. [Google Scholar]
- 4. Assari S, Burgard S. Black‐White differences in the effect of baseline depressive symptoms on deaths due to renal diseases: 25 year follow up of a nationally representative community sample. J Renal Inj Prev 2015; 4: 127–134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Navarro V. Race or class versus race and class: mortality differentials in the United States. Lancet 1990; 336: 1238–1240. [DOI] [PubMed] [Google Scholar]
- 6. Schulz AJ, Mullings L. sGender, Race, Class, and Health: Intersectional Approaches. San Francisco, CA: Jossey‐Bass, 2006. [Google Scholar]
- 7. Assari S. Distal, intermediate, and proximal mediators of racial disparities in renal disease mortality in the United States. J Nephropathol 2016; 5: 51–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Assari S. Number of Chronic Medical Conditions Fully Mediates the Effects of Race on Mortality; 25‐Year Follow‐Up of a Nationally Representative Sample of Americans. J Racial Ethn Health Disparities 2016. doi:10.1007/s40615‐016‐0266‐4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Cabassa LJ, Humensky J, Druss B, et al Do race, ethnicity, and psychiatric diagnoses matter in the prevalence of multiple chronic medical conditions? Med Care 2013; 51: 540–547. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Johnson‐Lawrence VD, Griffith DM, Watkins DC. The effects of race, ethnicity and mood/anxiety disorders on the chronic physical health conditions of men from a national sample. Am J Mens Health 2013; 7: 58S–67S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Lindhorst J, Alexander N, Blignaut J, et al Differences in hypertension between blacks and whites: an overview. Cardiovasc J Afr 2007; 18: 241–247. [PMC free article] [PubMed] [Google Scholar]
- 12. Signorello LB, Schlundt DG, Cohen SS, et al Comparing diabetes prevalence between African Americans and Whites of similar socioeconomic status. Am J Public Health 2007; 97: 2260–2267. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Jackson CL, Szklo M, Yeh HC, et al Black‐white disparities in overweight and obesity trends by educational attainment in the United States, 1997‐2008. J Obes 2013; 2013: 140743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Secrest AM, Costacou T, Gutelius B, et al Associations between socioeconomic status and major complications in type 1 diabetes: the Pittsburgh epidemiology of diabetes complication (EDC) Study. Ann Epidemiol 2011; 21: 374–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Baumann LC, Chang MW, Hoebeke R. Clinical outcomes for low‐income adults with hypertension and diabetes. Nurs Res 2002; 51: 191–198. [DOI] [PubMed] [Google Scholar]
- 16. Abate N, Chandalia M. The impact of ethnicity on type 2 diabetes. J Diabetes Complications 2003; 17: 39–58. [DOI] [PubMed] [Google Scholar]
- 17. Sampson UK, Edwards TL, Jahangir E, et al Factors associated with the prevalence of hypertension in the southeastern United States: insights from 69,211 blacks and whites in the Southern Community Cohort Study. Circ Cardiovasc Qual Outcomes 2014; 7: 33–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Laville M, Lengani A, Serme D, et al Epidemiological profile of hypertensive disease and renal risk factors in black Africa. J Hypertens 1994; 12: 839–843. [DOI] [PubMed] [Google Scholar]
- 19. Seedat YK. Hypertension in black South Africans. J Hum Hypertens 1999; 13: 96–103. [DOI] [PubMed] [Google Scholar]
- 20. Richardson AD, Piepho RW. Effect of race on hypertension and antihypertensive therapy. Int J Clin Pharmacol Ther 2000; 38: 75–79. [DOI] [PubMed] [Google Scholar]
- 21. House JS, Lepkowski JM, Kinney AM, et al The Social Stratification of Aging and Health. J Health Soc Behav 1994; 35: 213–234. [PubMed] [Google Scholar]
- 22. House JS, Kessler RC, Herzog AR. Age, Socioeconomic Status, and Health. Milbank Q 1990; 68: 383–411. [PubMed] [Google Scholar]
- 23. Guimarães JM, Chor D, Werneck GL, et al Association between self‐rated health and mortality: 10 years follow‐up to the Pró‐Saúde cohort study. BMC Public Health 2012; 12: 676. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Radloff LS. The CES‐D scale a self‐report depression scale for research in the general population. Appl Psychol Measurement 1977; 1: 385–401. [Google Scholar]
- 25. Zhang W, O'Brien N, Forrest JI, et al Validating a shortened depression scale (10 item CES‐D) among HIV‐positive people in British Columbia, Canada. PLoS ONE 2012; 7: e40793. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Houle JN. Depressive symptoms and all‐cause mortality in a nationally representative longitudinal study with time‐varying covariates. Psychosom Med 2013; 75: 297–304. [DOI] [PubMed] [Google Scholar]
- 27. Assari S. Additive Effects of Anxiety and Depression on Body Mass Index among Blacks: role of Ethnicity and Gender. Int Cardiovasc Res J 2014; 8: 44–51. [PMC free article] [PubMed] [Google Scholar]
- 28. Gavin AR, Rue T, Takeuchi D. Racial/ethnic differences in the association between obesity and major depressive disorder: findings from the Comprehensive Psychiatric Epidemiology Surveys. Public Health Rep 2010; 125: 698–708. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Simon GE, Von Korff M, Saunders K, et al Association between obesity and psychiatric disorders in the US adult population. Arch Gen Psychiatry 2006; 63: 824–830. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Taylor AW, Dal Grande E, Gill TK, et al How valid are self‐reported height and weight? A comparison between CATI self‐report and clinic measurements using a large cohort study. Aust N Z J Public Health 2006; 30: 238–246. [DOI] [PubMed] [Google Scholar]
- 31. Lantz PM, House JS, Lepkowski JM, et al Socioeconomic factors, health behaviors, and mortality: results from a nationally representative prospective study of US adults. JAMA 1998; 279: 1703–1708. [DOI] [PubMed] [Google Scholar]
- 32. Anderson RN, Miniño AM, Hoyert DL, et al Comparability of cause of death between ICD‐9 and ICD‐10: preliminary estimates. Natl Vital Stat Rep 2001; 49: 1–32. [PubMed] [Google Scholar]
- 33. Centers for Disease Control and Prevention, National Center for Health Statistics . Instruction manual, part 9. ICD‐10 cause‐of‐death lists for tabulating mortality statistics (updated October 2002 to include ICD codes for terrorism deaths for data year 2001 and WHO updates to ICD‐10 for data year 2003). Available from: http://www.cdc.gov/nchs/data/dvs/im9_2002.pdf.pdf Accessed December 2016.
- 34. Hertz RP, Unger AN, Cornell JA, et al Racial disparities in hypertension prevalence, awareness, and management. Arch Intern Med 2005; 165: 2098–2104. [DOI] [PubMed] [Google Scholar]
- 35. Zell JA, Rhee JM, Ziogas A, et al Race, socioeconomic status, treatment, and survival time among pancreatic cancer cases in California. Cancer Epidemiol Biomarkers Prev 2007; 16: 546–552. [DOI] [PubMed] [Google Scholar]
- 36. Heisler M, Smith DM, Hayward RA, et al Racial disparities in diabetes care processes, outcomes, and treatment intensity. Med Care 2003; 41: 1221–1232. [DOI] [PubMed] [Google Scholar]
- 37. Kim J, Artinyan A, Mailey B, et al An interaction of race and ethnicity with socioeconomic status in rectal cancer outcomes. Ann Surg 2011; 253: 647–654. [DOI] [PubMed] [Google Scholar]
- 38. O'Connell MJ, Kasprow WJ, Rosenheck RA. Differential impact of supported housing on selected subgroups of homeless veterans with substance abuse histories. Psychiatr Serv 2012; 63: 1195–1205. [DOI] [PubMed] [Google Scholar]
- 39. Hudson DL, Puterman E, Bibbins‐Domingo K, et al Race, life course socioeconomic position, racial discrimination, depressive symptoms and self‐rated health. Soc Sci Med 2013; 97: 7–14. [DOI] [PubMed] [Google Scholar]
- 40. LaVeist T, Pollack K, Thorpe R Jr, et al Place, not race: disparities dissipate in southwest Baltimore when blacks and whites live under similar conditions. Health Aff (Millwood) 2011; 30: 1880–1887. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Williams DR. Race, socioeconomic status, and health the added effects of racism and discrimination. Ann N Y Acad Sci 1999; 896: 173–188. [DOI] [PubMed] [Google Scholar]
- 42. Williams DR, Lavizzo‐Mourey L, Warren RC. The concept of race and health status in America. Public Health Rep 1994; 109: 26. [PMC free article] [PubMed] [Google Scholar]
- 43. Assari S. Ethnic and Gender Differences in Additive Effects of Socio‐economics, Psychiatric Disorders, and Subjective Religiosity on Suicidal Ideation among Blacks. Int J Prev Med 2015; 6: 53. eCollection 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Farmer MM, Ferraro KF. Are racial disparities in health conditional on socioeconomic status? Soc Sci Med 2005; 60: 191–204. [DOI] [PubMed] [Google Scholar]
- 45. Assari S, Lankarani MM. Race and Urbanity Alter the Protective Effect of Education but not Income on Mortality. Front Public Health. 2016; 4: 100. eCollection 2016. P [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Assari S. Combined Racial and Gender Differences in the Long‐Term Predictive Role of Education on Depressive Symptoms and Chronic Medical Conditions. J Racial Ethn Health Disparities. 2016. doi:10.1007/s40615‐016‐0239‐7. [DOI] [PubMed] [Google Scholar]
- 47. Norris K, Nissenson AR. Race, gender, and socioeconomic disparities in CKD in the United States. J Am Soc Nephrol 2008; 19: 1261–1270. [DOI] [PubMed] [Google Scholar]
- 48. Assari S, Lankarani MM. Chronic Medical Conditions and Negative Affect; Racial Variation in Reciprocal Associations Over Time. Front Psychiatry 2016; 7: 140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Assari S, Lankarani MM. Does Multi‐morbidity Mediate the Effect of Socioeconomics on Self‐rated Health? Cross‐country Differences Int J Prev Med 2015; 6: 85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Assari S. Cross‐Country Differences in the Additive Effects of Socioeconomics, Health Behaviors and Medical Comorbidities on Disability among Older Adults with Heart Disease. J Tehran Heart Cent 2015; 10: 24–33. [PMC free article] [PubMed] [Google Scholar]
- 51. Assari S. Cross‐country variation in additive effects of socio‐economics, health behaviors, and comorbidities on subjective health of patients with diabetes. J Diabetes Metab Disord 2014; 13: 36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52. Assari S, Lankarani RM, Lankarani MM. Cross‐country differences in the association between diabetes and disability. J Diabetes Metab Disord 2014; 13: 3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53. Drake KA, Galanter JM, Burchard EG. Race, ethnicity and social class and the complex etiologies of asthma. Pharmacogenomics 2008; 9: 453–462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Krause JS, Broderick LE, Saladin LK, et al Racial disparities in health outcomes after spinal cord injury: mediating effects of education and income. J Spinal Cord Med 2006; 29: 17–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55. Brown SA, Saunders LL, Krause JS. Racial disparities in depression and life satisfaction after spinal cord injury: a Mediational model. Top Spinal Cord Inj Rehabil 2012; 18: 232–240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56. Boyer GS, Templin DW, Goring WP, et al Discrepancies between patient recall and the medical record. Potential impact on diagnosis and clinical assessment of chronic disease. Arch Intern Med 1995; 155: 1868–1872. [PubMed] [Google Scholar]
- 57. Kochanek KD, Arias E, Anderson RN. How did cause of death contribute to racial differences in life expectancy in the United States in 2010. NCHS Data Brief 2013; 125: 1–8. [PubMed] [Google Scholar]