characterizes the study sample by multimorbidity categories in our study sample above 21 years of age, alive, with at least one of the cardiometabolic, musculoskeletal, and respiratory conditions in year 2009. Thirty-four percent of our study sample had cardiometabolic conditions and 25% had both cardiometabolic and musculoskeletal disease clusters; 4% had both cardiometabolic and respiratory disease clusters. However, only 7% of the study sample had all the three, cardiometabolic, musculoskeletal, and respiratory disease clusters.
Weighted percentages of chronic condition clusters by sample characteristics. Medical expenditure panel survey, 2009.
summarizes number and weighted percentages of individuals with polypharmacy by selected characteristics. Women compared to men were significantly more likely to be on polypharmacy (OR = 1.41, 95% CI = 1.27–1.56). Individuals in older age groups 40–49, 50–64, 65–69, 70–74, and 75 and older were also significantly more likely to be on polypharmacy compared to individuals in the age group 22–39. The odds ratios ranged from 2.03 to 7.70. There was also a positive and significant association between total outpatient visit quartiles and polypharmacy. Individuals who had visits in the upper quartile (4th quartile) were 17 times as likely as those with visits in the 1st quartile (OR = 16.77; 95% CI = 12.5–22.4).
Number and weighted percent with polypharmacy. Unadjusted odds ratios and 95% CI from logistic regression on polypharmacy. Medical Expenditure Panel Survey, 2009.
We present weighted percentage of individuals with polypharmacy among different multimorbidity categories in the left panel of . As seen from , the highest rates (64.1%) of polypharmacy were found in sample individuals with all three (cardiometabolic and respiratory and musculoskeletal) disease clusters. The next highest rates (41.2% and 41.8%) were observed among those with cardiometabolic and musculoskeletal disease clusters and among those with cardiometabolic and respiratory disease clusters. The lowest rates were found in those with only musculoskeletal (7.9%) and only respiratory clusters (7.2%).
Weighted percentage with polypharmacy. Unadjusted and adjusted odds ratio and 95% confidence intervals for chronic condition clusters. From logistic regressions on polypharmacy. Medical Expenditure Panel Survey, 2009.
Unadjusted logistic regressions and multivariable logistic regressions were used to examine the association between chronic condition clusters and polypharmacy. Odds ratios (OR) and AORs with their 95% confidence intervals for polypharmacy are presented in . Compared to individuals with all the three disease clusters (cardiometabolic, musculoskeletal, and respiratory), those with either one or two disease clusters were significantly less likely to receive polypharmacy. The unadjusted odds ratios ranged from 0.04 among those with respiratory conditions only to 0.40 among those with cardiometabolic and respiratory disease clusters.
We also examined the differences in the likelihood of polypharmacy between different single condition clusters. Compared to individuals with cardiometabolic disease cluster only, those with musculoskeletal cluster only and respiratory cluster only had lower odds ratios of reporting polypharmacy (OR = 0.38 and OR = 0.35, resp.). On the other hand, there were no significant differences in ORs between individuals with musculoskeletal conditions only and respiratory conditions only (OR = 0.91, 95% CI = (0.59, 1.39)).
When examining the differences in the likelihood of polypharmacy by two disease clusters, we found that individuals with both cardiometabolic and musculoskeletal clusters were more likely to report polypharmacy compared to those with both musculoskeletal and respiratory clusters (OR = 1.77). Similarly, individuals with both cardiometabolic and respiratory clusters were more likely to report polypharmacy (OR = 1.82) as compared to those with both musculoskeletal and respiratory clusters. Individuals with cardiometabolic and respiratory clusters did not significantly differ in the OR of polypharmacy from those with cardiometabolic and musculoskeletal clusters (OR = 1.03, 95% CI = 0.80, 1.32).
Multivariable logistic regressions to assess the association between multimorbidity and polypharmacy that controlled for gender, age, race/ethnicity, marital status, education, poverty status, health insurance, usual source of care, perceived physical and mental health, smoking status, BMI, and exercise revealed similar findings. For example, compared to individuals with cardiometabolic disease cluster only, those with musculoskeletal cluster only (AOR = 0.38; 95% CI = 0.30–0.49) and respiratory cluster only had lower likelihood (AOR = 0.62; 95% CI = 0.40–0.96) of polypharmacy. However, individuals with respiratory conditions had higher likelihood of receiving polypharmacy as compared to individuals with musculoskeletal conditions only (AOR = 1.62, 95% CI = 1.01–2.60).