Poster Paper:
Are Targeted Social Service Investments Associated with Better Health Outcomes for Older Adults? a State-Level Analysis
*Names in bold indicate Presenter
Methods: We conducted a retrospective longitudinal study of state-level spending per person 65 years or older and older adult health outcomes across all 50 states and the District of Columbia during 2007-2013 (n=357 state-years). We employed multivariable regression to estimate associations between state-level OAA expenditures (State Program Reports) and nine older adult health outcomes (Medicare Geographic Variation Public Use File) including arthritis, osteoporosis, hypertension, high cholesterol, diabetes, stroke, chronic obstructive pulmonary disorder, depression, and Alzheimer’s disease. All health outcomes were measured as the percent of Medicare beneficiaries 65+ in the state (enrolled in Part A and Part B) diagnosed with the chronic condition. OAA service expenditures were aggregated into three categories - nutritional, access-related, and in-home support services – and separate analyses were conducted for each service category. Spending variables (adjusted for inflation) were lagged by five years to better reflect the nature of their influence on health outcomes. Models were adjusted for state GDP, state-level Medicare spending and number of primary care providers, urbanicity, state politics, and socio-demographic factors of the 65+ state population (sex, age, race, and Medicaid eligibility of Medicare beneficiaries), as well as year and census region fixed effects. Robust standard errors, clustered by state, were also incorporated.
Principal Findings: Mean spending per person 65 years or older for nutritional, access-related, and in-home support services was $51.59, $18.08, and $10.76, respectively. Increased expenditures for all three service categories were significantly associated with lower prevalence of diabetes, hypertension, high cholesterol, and stroke. Increased expenditures for access-related and in-home services were also significantly associated with lower prevalence of heart attack. Significant coefficients for an additional ten dollar investment per older adult ranged from -0.04 to -0.47 (p<0.01) for nutritional services, -0.02 to -1.08 (p<0.05) for access-related services, and -0.01 to -0.26 (p<0.10) for in-home support services. Similar associations were found for the other chronic conditions, but did not reach statistical significance.
Conclusions: Our findings indicate that additional state capacity to provide social services may be protective for older adult health. Specifically, results from this study suggest that investments in non-medical services for older adults may be beneficial for cardiovascular conditions and diabetes among older populations in the states. Future work may elucidate the mechanisms by which these services influence health of older adults, as well as ways to optimize state-level social service investments.