Lifecourse socioeconomic circumstances and multimorbidity among older adults

Chronic conditions generally cluster and individuals with one chronic condition are likely to have other conditions as well. As such, many older adults now manage two or more chronic conditions at the same time (i.e. multimorbidity)[1, 2]; and evidence suggests that the presence of multimorbidity is expected to continue to rise [3‐5]. The influence of childhood material conditions and lifetime earnings on multimorbidity has yet to be investigated. Considering the impending shift in the US population demographic as the baby-boomers reach age 65, which will yield an elderly population characterized by declining death rates, increasing life expectancy, and increasing health care costs,[6, 7] an understanding of the correlates of multimorbidity in older adults is a crucial issue for population health research [8, 9].

Table 1 lists the descriptive statistics. The average age for the full sample was 65 years old. Nineteen percent of the sample reported childhood financial hardship. The mean number of chronic conditions was 1.24, with 30% (N = 2,193) reporting no chronic conditions. The mean number of chronic conditions for those reporting childhood financial hardship was 1.35 and the mean number of chronic conditions for those reporting no childhood financial hardship was 1.20. The median average annual lifetime earnings up to age 50 (CPI-U adjusted) was $19,410 with an inter-quartile range of $20,235 (see Table 1).

Bivariate analysis suggests differences in multimorbidity across all demographic and socioeconomic categories except Hispanic ethnicity (see Table 1). Unadjusted rate ratios showed that men reported slightly higher multimorbidity than women. In particular, the expected number of chronic conditions in men was 8% higher than in women. For non-white respondents, the unadjusted expected number of chronic conditions was 21% higher compared to white respondents. The unadjusted expected number of chronic conditions for those age 60-69 and 70 and older was 56% and 85%, respectively, greater than the respondents age 50-59. The unadjusted expected number of chronic conditions across education categories was 16% higher for those with educational attainment of high school/GED and 46% higher for those with less than high school compared to the some college or more group of respondents. No statistically significant differences in multimorbidity were noted in Hispanic ethnicity; therefore, ethnicity was not included in subsequent models (see Table 2).

This study investigated the association among childhood financial hardship, lifetime earnings, and multimorbidity in a sample of older adults. Our findings revealed that after controlling for socioeconomic and demographic characteristics, childhood financial hardship was positively associated with a higher number of the chronic conditions. This suggests that over and above the influence of age, race, and educational attainment, childhood financial hardship exerts an influence on the multimorbidity of the six chronic conditions measured in this study for older adults. Our findings also indicate that lifetime earnings was negatively associated with multimorbidity, although the noted association was relatively small. In particular, we showed that as the average annual income during young and middle adulthood increases by $10,000 the number of chronic conditions (as measured in this study) decreases by 5%. Additionally, when we included both childhood financial hardship and lifetime earnings in our models, the association between lifetime earnings and multimorbidity remained unchanged and the association between childhood financial hardship and multimorbidity was only slightly reduced. However, our tests of interactions revealed that lifetime earnings significantly modifies the relationship between childhood financial hardship and multimorbidity. This suggests that the influence of financial hardship in childhood on subsequent multimorbidity may be altered by earnings occurring in young and middle adulthood. More specifically, our findings showed that for older adults experiencing childhood financial hardship an increase by $10,000 in average annual earnings reduces the expected number of chronic conditions by 5%. Lastly, although we were primarily interested in determining the association among childhood financial hardship, lifetime earnings, and multimorbidity over and above the influence of other socioeconomic indicators such as educational attainment, it should be noted that increasing education was not consistently associated with an increase in the count of the six chronic conditions in this study. In particular, educational attainment was not associated with the absence of morbidity; and when compared to the less than high school group, only the high school/GED category showed a protective association with multimorbidity.

We evaluated our hypotheses using ZIP regression; and, the benefit of a ZIP modeling approach is the simultaneous estimation of factors associated with multimorbidity and the absence of morbidity. In fully adjusted models, childhood financial hardship was not associated with the absence of morbidity and lifetime earnings was only associated with the absence of morbidity in unadjusted models. Consistent with the literature, the logistic portion of our ZIP models showed that age was strongly negatively associated with the absence of morbidity [37]. Lastly, the widest confidence intervals were noted in the logistic portion of the childhood financial hardship ZIP models. This may indicate poor model fit and suggests that the potential socioeconomic and demographic factors associated with determining the number of diseases (of those measured in this study) might be substantively different from the factors associated with determining the absence of morbidity.

The lack of an association between the several indicators of socioeconomic circumstances in this study (i.e. childhood financial hardship, lifetime earnings, and educational attainment) and demographic characteristics (i.e. race and gender) and the absence of morbidity was unexpected. Research has shown a negative association between socioeconomic status and morbidity[38, 39] and multimorbidity[40, 41]; and studies on successful aging (where one component is the absence or low risk of morbidity[42]) have shown that childhood and mid-life socioeconomic circumstances[43], and the stability of financial resources[44] were positively associated with the absence of morbidity in older adults. However, the evidence on the association between socioeconomic factors and successful aging is equivocal[45‐47]. It has also been suggested that age effects on self-reported morbidity may overshadow socioeconomic effects[48]. Not surprising, our results show an independent association between the absence of morbidity and age, even when childhood financial hardship and lifetime earnings are included in the model. Demographic characteristics[48] and psychosocial and behavioral factors[49] across the lifecourse may prove more important than socioeconomic factors in determining the absence of morbidity; however, the influence of lifecourse socioeconomic factors on the absence of morbidity requires further exploration to fully elucidate their role in successful aging in general and the absence of morbidity in particular among older adults.

There are limitations to the present study. First, respondents in the HRS were asked to recall childhood financial hardship experiences; as a result, the measures of childhood financial hardship may be subject to recall bias. Second, our sample only included those individuals who provided permission for their social security records to be linked to their HRS records potentially introducing sample selection bias. However, it has been shown that the HRS respondents who grant permission for their social security were not very much different from those who did not [60]. In addition, the social security data linked to the HRS is capped at the taxable maximum for a given year and only includes earnings subject to Social Security deduction. Next, the literature remains inconsistent in defining multimorbidity [61, 62]; for example, some studies define multimorbidity as the count of two or more chronic conditions,[40] others use a combination of specific diseases, and still others use a simple count of chronic conditions reported or obtained from medical records [63, 64]. As a result, little guidance is provided in selecting a measure for multimorbidity and the specific diseases such a measure should contain. However, the chronic diseases selected for the present study are consistent with those of other studies of multimorbidity among older adults [21, 65]. Lastly, our study is cross-sectional so causal inferences cannot be definitively made regarding the associations noted.

Multimorbidity has considerable implications for health-related research, health care, and many government systems (e.g. Medicare, Medicaid, and Veterans Administration). Due to the single disease focus in research and health care[66, 67], we know relatively little about the correlates of multimorbidity [68, 69]. Yet, health care models on the delivery of clinical services for the elderly with multiple chronic conditions have recently been introduced highlighting the impact the growing problem of multimorbidity has on health care delivery [70, 71]. However, we know little about the factors that are associated with the various combinations of morbidity experienced by older adults, and such knowledge could potentially improve care and delay mortality for many managing multiple chronic conditions [72]. Additional research focused on the specific socioeconomic factors associated with multimorbidity across the lifecourse can also be used to inform the development of appropriate interventions that target socioeconomic groups at greatest risk for multimorbidity before they enter the health care system. In particular, interventions that target specific socioeconomic pathways might prove useful in helping reduce the burden of multimorbidity; more specifically, interventions that focus on material resources such as reducing hardships during childhood and increasing earnings throughout young and middle adulthood may have a substantial impact on prevention efforts[73] with those at risk for various combinations of chronic disease multimorbidity.