Elevated Depressive Symptoms and the Risk of Stroke among the Mexican Older Population

Abstract

Background/objectives: Several longitudinal studies in high-income countries suggest that depression increases stroke risk. However, few prior studies have evaluated this association in low- and middle-income countries (LMICs), where rapidly aging populations may have markedly different vascular risk profiles.

Design: Prospective cohort study.

Setting: The Mexican Health and Aging Study is a national population-based study of older adults in Mexico.

Participants: A total of 10,693 Mexican adults aged 50 and older enrolled in 2001 with no history of prior stroke.

Measurements: Depressive symptoms were assessed with a modified 9-item Centers for Epidemiologic Studies Depression Scale (elevated depressive symptom cutoff ≥5) in 2001 and 2003. We evaluated associations between baseline and short-term (2-year) changes in elevated depressive symptoms (categorized as stable low, recently remitted, recent-onset, or stable high symptoms) with incident self-reported or next-of-kin reported doctor-diagnosed stroke through 2015 using Cox proportional hazards models and sensitivity analyses applying inverse probability weights.

Results: Over an average follow-up of 11.4 years (standard deviation = 4.2), 10,693 respondents reported 546 incident strokes. Individuals with elevated baseline depressive symptoms experienced a moderately higher hazard of incident stroke (hazard ratio [HR] = 1.13; 95% confidence interval [CI] = .95-1.36) compared with those without elevated baseline depressive symptoms. In analyses of short-term changes in elevated depressive symptoms (n = 8,808; 414 incident stokes), participants with recent-onset (HR = 1.38; 95% CI = 1.06-1.81) or stable high (HR = 1.42; 95% CI = 1.10-1.84) elevated depressive symptoms had a greater hazard of incident stroke compared to those with stable low/no depressive symptoms, whereas recently remitted (HR = 1.01; 95% CI = .74-1.37) symptoms was not associated with stroke hazard.

Conclusion: Strategies to reduce depressive symptoms merit evaluation as approaches to prevent stroke in middle-income countries. Findings are similar to those in high-income countries but should be replicated in other LMICs.

Keywords: aging; depression; epidemiology; stroke.

Figure 1.

Hazard ratios (HRs) for elevated depressive symptoms (Center for Epidemiologic Studies Depression Scale [CES-D] score ≥ 5) at baseline (in 2001) and risk of incident stroke. Reference: low/no elevated depressive symptoms. Model 1: unadjusted; Model 2: adjusted for age and sex; Model 3: additionally adjusted for education, marital status, urbanicity, and state of residence; Model 4: inverse probability weighted Model 1 adjusted for age and sex (inverse probability of exposure weights account for education, marital status, urbanicity, state of residence, income, hypertension, and diabetes) (Supplementary Table S2 provides HR estimates).

Figure 2.

Hazard ratios (HRs) for short-term (2001–2003) patterns of elevated depressive symptoms (CES-D score ≥ 5) and risk of incident stroke. Reference: stable low/no elevated depressive symptoms. Model 1: unadjusted; Model 2: adjusted for age and sex; Model 3: additionally, adjusted for education, marital status, urbanicity, and state of residence; Model 4: inverse probability weighted Model 1 adjusted for age and sex (inverse probability of exposure weights account for education, marital status, urbanicity, state of residence, income, hypertension, and diabetes). (Supplementary Table S6 provides HR estimates).