Association of Cardiovascular Events With Spouse’s Subsequent Dementia

Abstract

This cohort study evaluates whether a spousal cardiovascular event is associated with one’s own risk of dementia.

Dementia is estimated to affect 152.8 million people worldwide in 2050.¹ Although studies have reported cardiovascular diseases (CVDs) as one of the major risk factors for dementia at the individual level, evidence of the association between CVD and dementia at the interpersonal level is lacking.² We evaluated whether a spousal CVD event is associated with one’s own dementia risk. Identifying how one spouse’s health is associated with the other spouse’s health can help direct timely, needed resources for people at risk of dementia.

Methods

This cohort study examined index individuals (primary insured) aged 65 years or older enrolled in the Japan Health Insurance Association insurance program. Index individuals were dementia free when their spouse (dependent) experienced incident CVD with an onset date between April 1, 2016, and March 31, 2022. To adjust for potential confounders and define time zero, we applied a target trial emulation framework.³ Each index individual was matched on age, sex, and income with a comparison person whose spouse had no CVD at the CVD onset date. Details on the insurance program, variable definitions, and matching process are in the eMethods in Supplement 1 and previous literature.⁴ After matching, 93 396 index individuals were analyzed (eFigure in Supplement 1). Kyoto University institutional review board approved this study with a waiver of informed consent because the data were retrospective. This study followed the STROBE reporting guideline.

After describing demographic characteristics, we drew the cumulative incidence curve and applied multivariable Poisson regression models to investigate the association between spouses’ CVD and individuals’ dementia. The model included age, sex, age of spouse, income, and disease histories of index individual and spouse (diabetes, hypertension, CVD, and dementia). To examine potential heterogeneity in associations between spouses’ CVD and new onset of individuals’ dementia by baseline characteristics, we conducted subgroup analyses by sex and histories of diabetes, hypertension, and CVD. Statistical analysis was performed with R, version 4.3.0. P values were from 2-sided tests with P < .05 indicating statistical significance.

Results

The index individuals (mean [SD] age, 68.8 [2.6] years) included 96.2% men and 3.8% women. Although the characteristics of index individuals were similar between exposed and unexposed groups, spouses in the exposed group exhibited a higher prevalence of disease histories than those in the unexposed group.

Over the median follow-up of 24 months (IQR, 10-42 months), 559 and 1653 events of dementia and death, respectively, occurred among index individuals. The exposed group had a higher cumulative incidence of dementia than the unexposed group (Figure 1). Poisson regression analysis showed the exposed group had a higher incidence rate ratio of dementia than the unexposed group (1.32 [95% CI, 1.10-1.57]); we found no evidence of heterogeneity by baseline characteristics (Figure 2).

Figure 1. Cumulative Incidence Curve for the Matched Index Individuals’ Incidence of Dementia by Spouses’ Cardiovascular Disease (CVD) Event.

A cumulative incidence function was applied to account for death of the index individuals.

Figure 2. Incidence Rate Ratios (IRRs) of Index Individuals’ Dementia Diagnosis Associated With Spouses’ Cardiovascular Disease (CVD) Events.

Each stratum was defined by characteristics of index individuals. The IRRs are computed via multivariable Poisson regression models adjusted for age, age of spouse, sex (not included in male and female), income, index individuals’ history of diabetes (not included in history of diabetes and no history of diabetes), hypertension (not included in history of hypertension and no history of hypertension), and CVD (not included in history of CVD and no history of CVD), as well as spouses’ history of diabetes, hypertension, and dementia.

Discussion

In this nationwide cohort in Japan, spousal CVD events were associated with increased dementia risk among individuals aged 65 years or older. This is one of the first and largest studies investigating potential interpersonal pathways associated with elevated dementia risk.

There are several possible mechanisms for our findings. First, traumatic occurrences can detrimentally affect cognitive function.⁵ Second, given the high mortality risk associated with CVD onset among older people, some could have encountered bereavement after their spouse’s CVD.⁶ Third, family members of patients with CVD might have changed their lifestyles (eg, diet and activity) after spouses’ CVD events, which could be associated with both CVD and dementia. Because caregiving burdens and social isolation could vary by sex, further research with enough female samples is needed to establish the sex-specific association.

People living with dementia need extra support to maintain a high quality of life, especially if their spouse is ill. Careful monitoring for dementia onset after spousal CVD events may help identify patients with emergent dementia. Limitations of this study include short follow-up, uncontrolled confounders, measurement errors, and lack of details regarding socioeconomic characteristics (eg, household income, educational attainment) and spouses’ deaths. The study also involves a predominantly male population, limiting generalizability.

Supplement 1.

eMethods. Details of JHIA Database, Variables, and Matching Process

eReferences.

eFigure. Sample Flow Chart

Supplement 2.

Data Sharing Statement