Independent and synergistic effects of pain, insomnia, and depression on falls among older adults: a longitudinal study

BACKGROUND: Few studies have examined the relationship between falls and pain, insomnia and depressive symptoms which are common and risk factors in older adults. We aimed to examine the independent and synergistic effects of these risk factors on future falls among older adults. METHODS: We used data of 2558 community-dwelling older adults from 2011 (Y1) to 2015 (Y5) of the National Health and Aging Trends Study (NHATS). Pain was determined by whether participants reported bothersome pain in the last month. Insomnia was assessed by two questions about how often the participants had trouble falling asleep and maintaining sleep. Depressive symptoms were assessed by Patient Health Questionnaire-2. Generalized estimation equation (GEE) models were used to examine the independent effects of pain, insomnia and depressive symptoms at prior-wave (period y-1) on falls at current wave (period y) adjusting for covariates (age, sex, education, race/ethnicity, living arrangement, BMI, smoking, vigorous activities, number of chronic illnesses and hospitalization). The significance of the three-way interaction of these factors (pain*insomnia*depression) was tested using the aforementioned GEE models to determine their synergistic effects on falls. RESULTS: Overall, the participants were mainly 65–79 years old (68%), female (57%) and non-Hispanic White (70%). At Y1, 50.0% of the participants reported pain, 22.6% reported insomnia and 9.9% reported depressive symptoms. The incidence of falls from Y2 to Y5 was 22.4, 26.0, 28.3, and 28.9%, respectively. Participants with pain (Odds ratio [OR], 95% confidence interval [CI] = 1.36, 1.23–1.50) and depressive symptoms (OR, 95% CI = 1.43, 1.23–1.67) had high rates of falling adjusting for covariates. After further adjustment for insomnia and depressive symptoms, pain independently predicted falls (OR, 95% CI = 1.36, 1.22–1.51). Depressive symptoms also independently predicted falls after further adjusting for pain and insomnia (OR, 95% CI = 1.40, 1.20–1.63). After adjusting for pain and depression, the independent effects of insomnia were not significant. None of the interaction terms of the three risk factors were significant, suggesting an absence of their synergistic effects. CONCLUSIONS: Pain and depressive symptoms independently predict falls, but synergistic effects seem absent. Further research is needed to develop effective strategies for reducing falls in older adults, particularly with pain and depressive symptoms.