Linear association between number of modifiable risk factors and multiple chronic conditions: Results from the Behavioral Risk Factor Surveillance System

Highlights

• We defined multiple (≥ 2) chronic conditions (MCCs) with 8 or 12 chronic conditions.

• We studied associations of MCCs with composite measures of up to 9 risk factors.

• Linear increase in MCC rates with each added risk factor, starting with one.

• All 9 risk factors are important.

• Results suggest potential for primary prevention of a wide range of conditions.

Abstract

Multiple (≥ 2) chronic conditions (MCCs) are responsible for a large fraction of healthcare costs. Our aim was to examine possible associations between MCCs and composite measures of behavioral risk factors (RFs). Data were publicly available 2013 Behavioral Risk Factor Surveillance System and included 483,865 non-institutionalized US adults ages ≥ 18 years. Chronic conditions included asthma, arthritis, chronic obstructive pulmonary disease, cognitive impairment, heart disease, stroke, cancer, and kidney disease. RFs included obesity, current smoking, sedentary lifestyle, inadequate fruit and vegetable consumption, and sleeping other than 7–8 h, while depression, hypertension, high cholesterol, and diabetes were considered in each category. Stata was used to study associations between 2 different MCCs and 2 composite measures of RFs in both unadjusted and adjusted analysis. Over 96% of respondents reported ≥ 1 of the 9 RFs and 71.5% reported ≥ 1 of the chronic conditions. For each combination there was a linear increase (with similar slopes) in MCC rate with more RFs and a statistically significant increase in adjusted odds ratios (ORs) for the MCC with each additional RF. For the MCC based on 8 chronic conditions, ORs were 1.3 (95% CI 1.1, 1.6) for 1 RF, 2.3 (1.9, 2.7) for 2, 3.7 (3.1, 4.4) for 3, 5.7 (4.8, 6.8) for 4, 9.1 (7.6, 10.8) for 5, 14.6 (12.2, 17.4) for 6, 24.0 (19.7, 29.2) for 7, 38.1 (29.6, 48.9) for 8, and 100.0 (56.3, 177.8) for all 9, each vs. zero RFs. Findings highlight the need for effective integrated programs to address multiple RFs and chronic conditions.

Introduction

Adults with multiple (≥ 2) chronic conditions (MCCs) account for about two-thirds of all healthcare costs in the U.S. (Anderson, 2010). MCCs are a major factor in the rise in Medicare spending (Thorpe et al., 2010), estimated to be responsible for 93% of those costs (Centers for Medicare and Medicaid Services, 2012). There is no standard definition of chronic conditions included in MCCs (Willadsen et al., 2016, Goodman et al., 2013) but chronic diseases, risk factors, mental health problems, and cognitive impairment can be among them (CMS: Centers for Medicare and Medicaid Services, n.d, U.S. Department of Health and Human Services, n.d). More attention is starting to be focused on MCCs as their contribution to health care costs is recognized (Centers for Medicare and Medicaid Services, 2012, Goodman et al., 2013, U.S. Department of Health and Human Services, n.d, Gupta, 2016). While a recent review (Willadsen et al., 2016) of 163 MCC studies found that 85% included risk factors (RFs) in their definitions of MCCs, apparently none studied possible associations between MCCs and RFs. Behavioral risk factors such as hypertension, obesity, and smoking have been shown to be associated with many separate chronic conditions (Brownson et al., 2010). Risk factors can also occur concurrently and are sometimes studied using composite measures. For example, Adams et al. (2016) and Liu et al. (2016) studied slightly different combinations of 5 RFs and both found that 92%–94% of all adults reported at least one RF. Addressing RFs collectively may help in understanding how they might predict MCCs, which in turn could inform clinical and public health practice.

Our objective for this current work was to study MCCs based on different definitions and their associations with composite measures of up to 9 risk factors. The chronic conditions chosen were asthma, arthritis, heart disease, stroke, chronic obstructive pulmonary disease (COPD), cognitive impairment, cancer other than skin, and chronic kidney disease. The RFs were current smoking, sedentary lifestyle, inadequate fruit and vegetable consumption, sleeping other than 7–8 h, and obesity. Because diabetes, hypertension, high cholesterol, and depression can be considered either chronic conditions (Willadsen et al., 2016, Goodman et al., 2013) or RFs (Brownson et al., 2010) they would be included in the study in each category. Prevalence rates of the composite measures plus their associations with each other would be studied. We would also test the hypothesis that there is a linear association between the number of RF's and MCC rates as was found for other outcomes (Adams and Grandpre, 2016). The hope was that results might aid in the development and targeting of integrated prevention programs addressing multiple RFs aimed at reducing rates of MCCs and lowering associated health care costs.