Diabetes-Multimorbidity Combinations and Disability Among Middle-aged and Older Adults

Background: Older adults with diabetes rarely have only one chronic disease. As a result, there is a need to re-conceptualize research and clinical practice to address the growing number of older Americans with diabetes and concurrent chronic diseases (diabetes-multimorbidity).

Objective: To identify prevalent multimorbidity combinations and examine their association with poor functional status among a nationally representative sample of middle-aged and older adults with diabetes.

Design: A prospective cohort study of the 2012-2014 Health and Retirement Study (HRS) data. We identified the most prevalent diabetes-multimorbidity combinations and estimated negative binomial models of diabetes-multimorbidity on prospective disability.

Participants: Analytic sample included 3841 HRS participants with diabetes, aged 51 years and older.

Main measures: The main outcome measure was the combined activities of daily living (ADL)-instrumental activities of daily living (IADL) index (range 0-11; higher index denotes higher disability). The main independent variables were diabetes-multimorbidity combination groups, defined as the co-occurrence of diabetes and at least one of six somatic chronic diseases (hypertension, cardiovascular disease, lung disease, cancer, arthritis, and stroke) and/or two mental chronic conditions (cognitive impairment and high depressive symptoms (CESD score ≥ 4).

Key results: The three most prevalent multimorbidity combinations were, in rank-order diabetes-arthritis-hypertension (n = 694, 18.1%); diabetes-hypertension (n = 481, 12.5%); and diabetes-arthritis-hypertension-heart disease (n = 383, 10%). Diabetes-multimorbidity combinations that included high depressive symptoms or stroke had significantly higher counts of ADL-IADL limitations compared with diabetes-only. In head-to-head comparisons of diabetes-multimorbidity combinations, high depressive symptoms or stroke added to somatic multimorbidity combinations was associated with a higher count of ADL-IADL limitations (diabetes-arthritis-hypertension-high depressive symptoms vs. diabetes-arthritis-hypertension: IRR = 1.95 [1.13, 3.38]; diabetes-arthritis-hypertension-stroke vs. diabetes-arthritis-hypertension: IRR = 2.09 [1.15, 3.82]) even after adjusting for age, gender, education, race/ethnicity, BMI, baseline ADL-IADL, and diabetes duration. Coefficients were robust to further adjustment for diabetes treatment.

Conclusions: Depressive symptoms or stroke added onto other multimorbidity combinations may pose a substantial functional burden for middle-aged and older adults with diabetes.