A random effect logistic regression model of major depressive disorder among ageing Nigerians

Abstract

Major Depressive Disorder (MDD) is a major public health problem in Nigeria and has severely devastating effects on the elderly. Previous studies on MDD among elderly Nigerians have utilized cross sectional designs which are descriptive in nature and have not investigated differences in setting and time-occurrence of MDD. Therefore this study employed a random effect logistic regression model to determine the relative effects/contributions of individual and environmental factors in the occurrence of MDD. A secondary analysis of a four-year longitudinal data from the Ibadan Study of Ageing was conducted. A total of 2,149 elderly Nigerians participated in the study between 2003 and 2009. The Geriatric Depression Scale was used to assess MDD and consequently classified as “present” for scores ranging from 10 to 30 and “absent” for scores ranging from 0 to 9. A random effect logistic regression model was fitted to determine factors predicting MDD. Odds ratios (OR), 95% confidence intervals, and Intra-class Correlation Coefficients (ICC) for each random effect was estimated. The overall prevalence of MDD was 27.28%. Significant predictors of MDD included “no-contact with family members” (OR=2.9, 95%CI: 1.26-6.70), “no-contact with friends” (OR=1.32, 95%CI: 1.05-1.67)), non-participation in family activities (OR=2.07, 95%CI: 1.63-2.43), non-participation in community activities (OR=1.93, 95%CI: 1.54-2.43), and good quality of health (OR=0.25, 95%CI: 0.15-0.27). Disparities in the occurrence of MDD among the elderly were attributable to enumeration areas (6%) and the individuals (22%). Social isolation factors and self-reported quality of health are significant predictors of MDD among elderly Nigerians.