Do stroke and hypertension mediate associations between exposure to air pollutants and dementia?

By Pooja Toshniwal PahariaReviewed by Lily Ramsey, LLMSep 25 2023

In a recent study published in JAMA Network Open, researchers investigated whether stroke or hypertension modifies or mediates the relationship between particulate matter exposure (PM_2.5) in the air and new-onset dementia in the United States.

Study: Hypertension and Stroke as Mediators of Air Pollution Exposure and Incident Dementia. Image Credit: pikselstock/Shutterstock.com

Background

The cognitive decline among individuals with dementia affects their abilities to perform routine activities and increases dependency on others. PM_2.5 exposure has shown constant associations with cardiovascular illness, which may increase dementia risk.

PM_2.5 may mediate the dementia risk through vascular dysfunction; however, few prior studies have investigated this potential underlying mechanism.

About the study

In the present study, researchers investigated the potential modifying or mediating effects of stroke and hypertension on the link between PM2.5 exposure and new-onset dementia.

The study was conducted as a subset of the Environmental Predictors of Cognitive Health and Aging (EPOCH) study, using biennial survey information obtained from 1998 to 2016 from the Health and Retirement Study (HRS) respondents.

Eligible individuals were aged above 50 with no history of dementia and had adequate data on mediators, exposures, outcomes, and demographics from the Health and Retirement Study.

Individuals with missing information on exposure, main covariates, or mediators were excluded from the analysis. The data were analyzed between August and November 2022. The study exposure was PM_2.5 exposure, calculated over the previous ten years for each individual using spatiotemporal models and residential histories.

The models incorporated measurements from the United States Environmental Protection Agency (US EPA) regulatory networks and various research studies, with over 300 geographical covariates characterizing land cover and usage, transportation, emission sources, vegetation, and population density.

The outcome measure was new-onset dementia, identified using validated algorithms based on informant reports and cognitive testing.

A decomposition-type causal mediation evaluation was performed to quantitatively assess the extent to which stroke and hypertension modified or mediated the link between PM_2.5 and new-onset dementia after adjusting for area-level and individual-level covariates.

Cox proportional hazards modeling and logistic regression modeling were performed to calculate the hazard ratios (HRs) and odds ratios (ORs), respectively.

Covariates included age, gender, body mass index (BMI), race, ethnicity, education, income, smoking status, urbanicity, drinking status, and socioeconomic status.

In addition, the interquartile range (IQR) values were determined. Individuals were identified as hypertensive or having a stroke if the participants themselves or proxies reported a doctor’s diagnosis of the conditions at follow-up.

Secondary analyses were performed by characterizing PM_2.5 from open fires and agriculture by multiplying PM_2.5 with spatially varying fractions of PM_2.5 from each source. 

Sensitivity analyses were also performed by stratifying mediation analyses by age and restricting the analysis to periods with high-quality PM_2.5 estimates (i.e., from 2006 onward).

Furthermore, the analyses were rerun using new-onset stroke as the mediator to ensure temporality between the study exposure and the mediator.

Results

Among 27,857 individuals, the researchers found no evidence that stroke or hypertension acted as modifiers or mediators of the relationship between PM_2.5 exposure and new-onset dementia.

The mean age of the participants was 61 years; 15,747 were female (57%); 19,249 were non-Hispanic Whites (69%); and 4,105 (15%) were diagnosed with dementia at follow-up (mean duration of 10 years).

Among individuals with dementia, 2,204 (54%) were hypertensive at study initiation, and 386 (nine percent) were diagnosed with hypertension at follow-up. 378 individuals (nine percent) had a stroke history at study initiation, while 673 (16%) received a stroke diagnosis at follow-up.

The interquartile range of baseline exposure to PM_2.5 was 11 to 15 μg per cubic millimeter. In the adjusted analysis, increased PM_2.5 exposure (per IQR) was unrelated to increased dementia risk (HR, 1.0).

The team observed positive relationships for stroke and hypertension with new-onset dementia, with HR values of 1.7 and 1.2, respectively, compared to individuals free of hypertension or stroke at follow-up.

However, there were no significant associations between PM_2.5 and stroke (OR per IQR increase in particulate matter exposure, 1.1) and no significant evidence for a relationship between PM_2.5 and hypertension (OR per IQR increase in particulate matter exposure, 1.0).

No evidence was found that stroke or hypertension acted as modifiers or mediators of the link between PM_2.5 exposure and new-onset dementia.

The non-mediated interactions between hypertension and PM_2.5 exposure contributed 39% to the excess relationship; however, the findings did not attain statistical significance.

The secondary analyses yielded similar results for PM_2.5 exposure from open fires as the total PM_2.5. Consistent results were obtained in the sensitivity analyses by restricting the comments to younger individuals and periods with high-quality exposure information. In addition, using new-onset stroke as the association mediator also did not alter the results.

Conclusion

Overall, the study findings showed that stroke or hypertension did not modify or mediate the link between PM_2.5 exposure in the air and incident dementia. However, hypertension may increase individual susceptibility to air pollution. Further research must be conducted to explore other modifiers and pathways.