Age at cardiovascular disease onset, dementia risk, and the role of lifestyle factors

Abstract INTRODUCTION We first examined the role of age at cardiovascular disease (CVD) onset for incident dementia, and then examined whether lifestyle factors at guideline‐recommended levels in individuals with CVD mitigates dementia risk. METHODS We used population‐based data (Whitehall II: n = 10,308/baseline 1985–1988/examinations every 4–5 years). Lifestyle factors (non‐smoking, body mass index [BMI], physical activity, diet) were extracted post‐CVD. RESULTS Over a median of 31.6 years, 3275 (32.1%) developed CVD. At age 70, risk of dementia was higher in individuals with CVD onset before (hazard ratio [HR] of incident dementia for participants with CVD before age 60, using participants without CVD at age 70 as the reference: 1.56, 95% confidence interal [CI] 1.18–2.08) but not after 60 years. In participants with CVD, a greater number of lifestyle factors at recommended levels post‐CVD was associated with a lower dementia risk (per lifestyle factor at recommended level HR: 0.73, 95% CI 0.59–0.92). DISCUSSION Our results suggest that early onset CVD is associated with a higher dementia risk at older ages. In those with CVD, the dementia risk was lower if lifestyle factors are at recommended levels following CVD diagnosis. Highlights CVD in midlife but not in late life is associated with a higher risk of dementia. Dementia risk in CVD patients is lower if their lifestyle factors are at recommended levels. These findings provide evidence to promote CVD prevention in midlife or earlier. Study findings also show the importance of a healthy lifestyle in those with CVD.

‡ Model 1 includes age (as the time scale), sex, race/ethnicity, education, marital status, birth cohort (5-year groups), and time between diagnosis of CVD and measure of lifestyle factors.

Supplementary Table 1. Characteristics of participants with CVD, overall and according to dementia status
NOTE.Data are means (SD) or median (25 th percentile, 75 th percentile).* CVD was defined as coronary heart disease, stroke, atrial fibrillation, or heart failure.

Supplementary Table 2. Characteristics of included participants with CVD, and those excluded due to missing data on lifestyle factors. All participants with CVD * (n=1,536) Participants with CVD * excluded due to missing data on lifestyle factors (n=391) Characteristics following onset of cardiovascular disease
3, 19.7) years.‡ Lifestyle factors measured after diagnosis of CVD (median delay between CVD and lifestyle factors measurement: 2.3 years (1.5; 3.7).§ Dietary habits at recommended level was defined as consumption of fruit and vegetable twice a day or more and consumption of high fibre bread.Abbreviations: CVD, cardiovascular disease.

Lifestyle factors following onset of cardiovascular disease ‡
NOTE.Data are means (SD) or median (25 th percentile, 75 th percentile).*CVDwasdefined as coronary heart disease, stroke, atrial fibrillation, or heart failure.†Dietaryhabits at recommended level was defined as consumption of fruit and vegetable twice a day or more and consumption of high fibre bread.Abbreviations: CVD, cardiovascular disease.Supplementary Table3.

Association between the lifestyle factor * score and incidence of dementia † in participants with CVD -consecutive exclusion of lifestyle factors
Lifestyle factors measured after diagnosis of CVD; body mass index at recommended level if <30 kg/m2; smoking at recommended level if not-smoker; physical activity at recommended level if ≥150 min/week of MPA OR ≥75 min/week of VPA OR ≥150 min/week of MVPA; diet at recommended level if fruit and vegetable consumption ≥twice/day AND high fibre bread consumption. *

Association between lifestyle factor * and incidence of dementia † in participants with CVD -excluding participants with prevalent diabetes ‡ (n=133)
‡ Model 1 includes age (as the time scale), sex, race/ethnicity, education, marital status, birth cohort (5-year groups), and time between diagnosis of CVD and measure of lifestyle factors.§Model 2 includes variables in model 1 + hypertension and diabetes.Abbreviations: CVD, cardiovascular disease; CI, confidence interval; MPA, moderate physical activity; VPA, vigorous physical activity; MVPA, moderate and vigorous physical activity.Supplementary Table 4.* Lifestyle factors measured after diagnosis of CVD; body mass index at recommended level if <30 kg/m 2 ; smoking at recommended level if not-smoker; physical activity at recommended level if ≥150 min/week of MPA OR ≥75 min/week of VPA OR ≥150 week of MVPA; diet if consumption of fruit and vegetable twice a day or more AND consumption of high fibre bread.†Median (25 th percentile, 75 th percentile) overall follow-up was 11.1 (6.6, 20.1) years, for participants without dementia it was 11.1 (6.6, 20.1) years and for participants with incident dementia it was 11.8 (5.6, 20.0) years at end of follow-up.‡Prevalent diabetes was defined as type 2 diabetes prior to or at CVD diagnosis.§Model 1 includes age (as the time scale), sex, race/ethnicity, education, marital status, birth cohort (5-year groups), and time between diagnosis of CVD and measure of lifestyle factors.¶Model 2 includes variables in model 1 + hypertension and diabetes.#Model 3 includes variables in model 2 + mutual adjustment for lifestyle factors.Abbreviations: CVD, cardiovascular disease; CI, confidence interval; MPA, moderate physical activity; VPA, vigorous physical activity; MVPA, moderate and vigorous physical activity; N/A, not applicable.Supplementary Table 5.

Association between lifestyle factors * and incidence of dementia † in participants with CVD -excluding participants with lifestyle factors assessed more than 6 years after CVD diagnosis (n=114)
* Lifestyle factors measured after diagnosis of CVD; body mass index at recommended level if <30 kg/m 2 ; smoking at recommended level if not-smoker; physical activity at recommended level if ≥150 min/week of MPA OR ≥75 min/week of VPA OR ≥150 week of MVPA; diet if consumption of fruit and vegetable twice a day or more AND consumption of high fibre bread.

Association between lifestyle factors * and incidence of dementia † in participants with CVD -restricting analysis to participants of white ethnicity (n=1,354)
Lifestyle factors measured after diagnosis of CVD; body mass index at recommended level if <30 kg/m 2 ; smoking at recommended level if not-smoker; physical activity at recommended level if ≥150 min/week of MPA OR ≥75 min/week of VPA OR ≥150 week of MVPA; diet if consumption of fruit and vegetable twice a day or more AND consumption of high fibre bread.
§ Model 2 includes variables in model 1 + hypertension and diabetes.¶Model3includes variables in model 2 + mutual adjustment for lifestyle factors.Abbreviations: CVD, cardiovascular disease; CI, confidence interval; MPA, moderate physical activity; VPA, vigorous physical activity; MVPA, moderate and vigorous physical activity; N/A, not applicable.Supplementary Table 7.* ‡ Model 1 includes age (as the time scale), sex, education, marital status, birth cohort (5-year groups), and time between diagnosis of CVD and measure of lifestyle factors.§Model 2 includes variables in model 1 + hypertension and diabetes.¶Model 3 includes variables in model 2 + mutual adjustment for lifestyle factors.Abbreviations: CVD, cardiovascular disease; CI, confidence interval; MPA, moderate physical activity; VPA, vigorous physical activity; MVPA, moderate and vigorous physical activity; N/A, not applicable.Supplementary

Table 8 . Association of lifestyle factors * and incidence of dementia in participants with CVD -using inverse-probability weighting to account for missing data
Lifestyle factors measured after diagnosis of CVD; body mass index at recommended level if <30 kg/m 2 ; smoking at recommended level if not-smoker; physical activity at recommended level if ≥150 min/week of MPA OR ≥75 min/week of VPA OR ≥150 min/week of MVPA; diet at recommended level if fruit and vegetable consumption ≥twice/day AND high fibre bread consumption.† Model 1 includes age (as the time scale), sex, education, marital status, birth cohort (5-year groups), and time between diagnosis of CVD and measure of lifestyle factors.Model 3 includes variables in model 2 + mutual adjustment for lifestyle factors.Abbreviations: CVD, cardiovascular disease; CI, confidence interval; MPA, moderate physical activity; VPA, vigorous physical activity; MVPA, moderate and vigorous physical activity; N/A, not applicable.