Psychological Distress and Heart Disease Mortality in the United States: Results from the 1997-2014 NHIS-NDI Record Linkage Study

2.1. Data

The data for this study are derived from the National Health Interview Survey (NHIS) linked to the National Death Index (NDI).20 As a nationally representative, annual cross-sectional household interview survey, NHIS provides demographic, socio-economic, and health characteristics of the civilian, non-institutionalized population in the US. The National Center for Health Statistics (NCHS) developed public-use versions of NHIS linked with death certificate records from the NDI. For this study, we used the 1997-2014 public-use linked mortality file containing 18 years of mortality follow-up data from the date of survey participation through December 31, 2014.21

2.2. Sample

The study sample was restricted to adults aged 18 and older from the years of 1997 to 2014 NHIS sample adult files. The sample ineligible for mortality follow-up was eliminated from the analysis. The study sample was defined as adults without missing values in covariates using the STATA “svy, subpop” option. Thereby, all missing values in the study sample were excluded from the analyses but used for weights. The final pooled eligible sample size excluding missing values was 513,081. For missing values for poverty status (16.2%), body mass index (BMI) (3.5%) and alcohol consumption (1.7%), we created missing covariate categories to prevent omission of many observations from the analysis.

2.3. Measurement

We examined two mortality-related outcomes: age-adjusted mortality rates and relative risk of heart disease mortality as measured by hazard ratios. Follow-up time for individuals who died during the study period was estimated by the number of months from the month/year of interview to the month/year of death. Since the NHIS-NDI database provides only the quarter of death, we assumed that death occurred in the middle of quarter, February, May, August, or November.22

2.4. Psychological Distress

Psychological distress was measured by the Kessler 6 (K6) nonspecific distress scale23 of six symptoms. Respondents were asked: “during the past 30 days, how often you feel (1) so sad that nothing could cheer you up; (2) nervous; (3) restless or fidgety; (4) hopeless; (5) that everything was an effort; and 6) worthless.” The scores [5-point scale: all (4); most (3); some (2); a little (1); none of the time (0)] of each of the six symptoms of mental illness or nonspecific psychological distress were summed to create the K6 scale, ranging in value from 0 to 24. We created a 5-level categorical variable with scores of 0, 1-2, 3-5, 6-12, and 13-24 from the K6 scale; a score of 13 or higher was defined as serious psychological distress (SPD).24

2.5. Covariates

Based on the previous literature, we selected the following covariates for model estimation: age, gender, race/ethnicity, nativity/immigrant status, marital status, education, poverty status, housing tenure, region of residence, self-reported health status, activity limitation, BMI, smoking status, alcohol consumption, and survey years.16,22,24–27 Age was categorized into 14 five-year age groups. Race/ethnicity consisted of non-Hispanic Whites, non-Hispanic Blacks, Hispanics, AIANs, APIs, and non-Hispanic other races. Nativity/immigrant status was categorized as US-born, who were born in one of the 50 states or the District of Columbia, or foreign-born. Educational attainment was defined by four categories as less than high school diploma, high school diploma, some college, and college degree or more. Marital status was categorized as currently married, widowed, divorced/separated, and never married. Poverty status was defined by five categories based on the ratio of family income to poverty threshold (<100%; >=100% & <200%; >=200% & <400%; >=400%; missing). Housing tenure was dichotomized, with 1 being renters and zero equaling homeowners. Region of residence was defined by four categories, Northeast, Midwest, South, and West. Self-reported health status was dichotomized, with 1 being fair or poor health and zero being excellent, very good, or good health. Activity limitation was defined as one for a person with any Instrumental Activities of Daily Living (IADL), Activities of Daily Living (ADL), functional, or activity limitations in any of the pertinent rounds, and zero otherwise. BMI was defined by four categories, <25, 25-29, 30-39, ≥40, and missing. Smoking status was defined by three categories as never, former, and current smokers. Alcohol consumption was defined by four categories: lifetime abstainer, former drinker, current drinker, and unknown.

2.6. Analytic Approach

We computed age-adjusted mortality rates per 100,000 person-years of exposure for heart disease according to psychological distress level. In computing person-years for age-specific and age-adjusted heart disease mortality rates, deaths were assumed to occur uniformly throughout the follow-up period rather than being clustered in the first part of the period. Mortality rates were age adjusted by the direct method using the age distribution of the 2000 US Standard Population.2 Standard errors of age-adjusted mortality rates were computed, and mortality rate ratios (RR) associated with psychological distress level were tested for statistical significance.

Cox proportional hazards models were used to derive relative risks of heart disease mortality, controlling for individual characteristics and year-fixed effects. The model assumes that hazard rates are a log-linear function of parameters representing the effects of covariates.28,29 Individuals surviving beyond the follow-up period and those dying from causes other than heart disease were treated as right-censored observations. The Cox models were estimated separately for females and males and for different racial/ethnic groups.

We checked the hazards proportionality assumption29 by inspecting the plots of ln(-ln{S(t)]) [log(-log) survival function] against survival time t for the various covariate categories including those for psychological distress, gender, race/ethnicity, education, poverty level, activity limitation, and self-reported health status.30 These plots were found to be approximately parallel and hence the proportionality assumption was taken to be satisfied by the data.

Complex survey design procedures were used to account for clustering, multiple stages of selection, and disproportionate sampling. To correct the bias from the ineligible adults for linkage to the NDI due to insufficient identifying data, we used eligibility-adjusted weights developed by NCHS, instead of the standard sample weight.31 The sample weights were adjusted by dividing by the number of pooling years. All the analyses were conducted by STATA 15 and the Cox model was fitted using stcox procedure.32

3.1. Descriptive Statistics

Table 1 provides descriptive statistics for individual characteristics by psychological distress levels. Approximately 3.1% of the sample had SPD (K6 score ≥13). The study sample consisted of non-Hispanic Whites (70.9%), non-Hispanic Blacks (11.4%), Hispanics (12.7%), and APIs (4.0%). The total number of heart disease deaths during the 18-year follow-up was 10,889.

3.2. Heart Disease Mortality Rates by Psychological Distress Levels

The age-adjusted heart disease mortality rate for adults with SPD was 366 deaths per 100,000 person-years, 75% higher than the rate of 209 deaths for those without psychological distress (Table 2). The age-adjusted mortality rate was 45% higher for those with K6 scores of 6-12 and 22% higher for those with K6 of 3-5, compared to that for adults with a K6 score of zero (no psychological distress). Adults with higher psychological distress levels (K6 scores) had consistently higher age-adjusted heart disease mortality rates. Females had 40% lower heart disease mortality than males (184 deaths vs. 305 deaths per 100,000 person-years). Males with SPD had a heart disease mortality rate of 536 deaths per 100,000 person-years, while females with SPD had a heart disease mortality rate of 286 (Table 2).

3.3. Cox Proportional Hazards Models

In the Cox regression model 1, controlling for age and survey year, the heart disease mortality risk was 121% higher (hazard ratio [HR]=2.21; 95% CI=1.99,2.45) in adults with SPD (p<0.001), compared with adult without psychological distress (Table 3). In model 2, after controlling for age and survey year and socioeconomic and demographic characteristics, the heart disease mortality risk was 96% higher (HR=1.96; 95% CI=1.77,2.18) in adults with SPD (p<.001), compared with adult without psychological distress. In model 3, after controlling for all covariates including age, survey year, sex, race/ethnicity, nativity/immigrant status, education, marital status, poverty status, housing tenure, region, self-assessed health status, activity limitation, BMI, smoking status, and alcohol consumption, the overall heart disease mortality risk was 14% higher (HR=1.14; 95% CI=1.02,1.28) in adults with SPD (p<.05) than that in adults without SPD. In all three models, adults with K6 scores of 3-5 or 6-12 experienced a significantly higher heart disease mortality risk than those without any psychological distress. In general, the higher the psychological distress level, the higher the heart disease mortality risk. Trend in mortality risks by psychological distress level was statistically significant for all models.

Table 4 provides heart disease mortality risks separately for males and females by psychological distress levels after adjusting for socioeconomic and demographic characteristics. For males, the heart disease mortality risk was 179% higher (HR=2.79; 95% CI=2.39,3.26) in the age-year-adjusted model, 113% higher (HR=2.13; 95% CI=1.82,2.49) in the sociodemographically-adjusted model, and 25% higher (HR=1.25; 95% CI=1.06,1.47) in the fully-adjusted model among adults with SPD, compared with adults without psychological distress. For females, the heart disease mortality risk was 121% higher (HR=2.21; 95% CI=1.90,2.56) in the age-year-adjusted model, and 81% higher (HR=1.81; 95% CI=1.56,2.10) in the sociodemographically-adjusted model among adults with SPD, and not significantly different in the fully-adjusted model, compared with adults without psychological distress.

Table 4 also shows differential effects of psychological distress on heart disease morality for the three largest racial/ethnic groups. For non-Hispanic Whites and non-Hispanic Blacks, higher levels of psychological distress were generally associated with higher age-adjusted mortality risks. After controlling for age and survey year, non-Hispanic Whites with SPD had 142% higher heart disease mortality risks (HR=2.42; 95% CI=2.13,2.76), while non-Hispanic Blacks with SPD had 76% higher heart disease mortality risks (HR=1.76; 95% CI=1.37,2.25), compared with their counterparts with no psychological distress. Hispanics with SPD had 39% higher heart disease mortality risks (HR=1.39; 95% CI=1.02,1.90), but there was no consistent dose-response relationship. After controlling for socioeconomic and demographic characteristics, non-Hispanic Whites with SPD had 113% higher heart disease mortality risks (HR=2.13; 95% CI=1.88,2.42), while non-Hispanic Blacks with SPD had 52% higher (HR=1.52; 95% CI=1.18,1.94) and Hispanics with SPD had 54% higher heart disease mortality risks (HR=1.54; 95% CI=1.12,2.10), compared with their counterparts with no psychological distress. We did not find statistically significant differences in the fully-adjusted model for non-Hispanic Blacks and Hispanics.

Figure 1 provides estimated survivor functions for males and females and non-Hispanic Whites, non-Hispanic Blacks, and Hispanics according to psychological distress levels after adjusting for socioeconomic and demographic covariates in the Cox models. The estimated survival functions were steeper for males than for females. At the end of the 8-year follow-up, 98.54% of males with SPD were expected to survive from heart disease, compared with 99.31% of males with no psychological distress; the corresponding survival rates for males were 96.00% and 98.10% at the end of the 16-year follow-up. At the end of the 8-year follow-up, 99.57% of females with SPD were expected to survive from heart disease, compared with 99.76% of females with no psychological distress; the corresponding survival rates for females were 98.74% and 99.30% at the end of the 16-year follow-up.

Close

Figure 1

(a-e). Adjusted Survivorship of US Adults Aged ≥18 Years by Psychological Distress Level (K6 Scores) and Sex or Race/ethnicity, 1997-2014 (derived from estimated Cox models of heart disease mortality adjusted for socioeconomic and demographic covariates)

Export to PPT

The estimated survival functions were slightly steeper for non-Hispanic Whites than for non-Hispanic Blacks or Hispanics. At the end of the 8-year follow-up, 99.23% of non-Hispanic Whites with SPD were expected to survive from heart disease, compared with 99.64% of non-Hispanic Whites with no psychological distress; the corresponding survival rates for non-Hispanic whites were 97.78% and 98.95% at the end of the 16-year follow-up. At the end of the 8-year follow-up, 99.07% of non-Hispanic Blacks with SPD were expected to survive from heart disease, compared with 99.39% of non-Hispanic Blacks with no psychological distress; the corresponding survival rates for non-Hispanic Blacks were 97.77% and 98.52% at the end of the 16-year follow-up. At the end of the 8-year follow-up, 99.37% of Hispanics with SPD were expected to survive from heart disease, compared with 99.59% of Hispanics with no psychological distress; the corresponding survival rates for Hispanics were 98.21% and 98.83% at the end of the 16-year follow-up.

We conducted a sensitivity analysis to check for reverse causality between psychological distress and heart disease mortality by excluding individuals with a history of heart disease at the time of the survey. In samples with and without a history of heart disease diagnosis at baseline, the association between psychological distress and heart disease mortality was similar (71% and 54% higher relative mortality risks respectively for those with SPD in the sociodemographically-adjusted models, which reduced to statistical non-significance after additional adjustment for health status and health-risk factors) (Table 5). We also performed a sensitivity analysis to examine temporal robustness of psychological distress by re-estimating models using 2-year, 5-year, and 10-year follow-up times. As shown in Table 6, the longer follow-up dilutes the mortality impact of baseline psychological distress, as psychological distress is expected to differ from its baseline level over the longer follow-up.

4.1. Limitations

This study has limitations. First, our study only contains the NHIS sample eligible for linkage to the NDI. Excluding samples ineligible for linkage may lead to biased mortality estimates. To address this bias, we used the adjusted original sampling weight to account for the NHIS-NDI mismatches.31 Second, our findings may be affected by the omitted-variable bias. While our Cox regression models were controlled for self-reported health status, BMI, activity limitation, and physical health status, there could be other potential confounders. Antipsychotic medication use,25 undiagnosed or diagnosed mental illness16 could affect the association between psychological distress and mortality. Third, symptoms in K6, such as “sad, nervous, restless, hopeless, effort, and worthless”, might not capture all diagnostic criteria of psychological distress. Although the 6 items capture several diagnostic criteria in Diagnostic and Statistical Manual of Mental Disorders (DSM-5) or ICD-10, the scale does not provide a clinical diagnosis of depression or anxiety.16 However, K6 is a valid measure of psychological distress in terms of good precision in the scale distribution and consistent levels of severity across socio-demographic subsamples.34 Fourth, since the NHIS excludes the institutionalized population, who may have higher levels of psychological distress and mortality, heart disease mortality risks associated with psychological distress in this study might be underestimated. Fifth, all the covariates in the NHIS-NDI database were time-fixed at the baseline as of the survey date. Several of the covariates such as SES, health status, behavioral risk factors, and psychological distress could have varied over the long mortality follow-up period of 18 years, which would have influenced their estimated impact on heart disease mortality. Future studies need to evaluate the temporal robustness of psychological distress patterns in heart disease mortality using longitudinal datasets with time-varying covariates.