UCLA

Introduction: Dementia is a progressive condition marked by a decline in cognitive abilities, impairing an individual’s daily functioning and leading to dependence. Currently, there are over 55 million dementia cases globally, with over 60% of cases concentrated in low- and middle-income countries (LMICs). The population in LMICs is particularly vulnerable for risk of dementia from exposure to indoor air pollutants due to highly prevalent use of biomass fuel for cooking and heating. Additionally in the United States, nearly half of the 16.5 million living military veterans are men 65 years or older who are vulnerable for dementia due to various hazardous exposures during their time in service. With an aging global population, it is imperative to focus on preventing and delaying the onset of dementia, including among individuals in LMICs and former US service members. This dissertation aimed to further characterize the risk of impaired cognitive function by exposure to indoor air pollutants and veteran status among marginalized populations in India and the US, respectively.Method: Using data obtained from a nationally representative cross-sectional survey in India, I estimated the effect of exposure to reported use of unclean cooking fuel compared with clean cooking fuel on culturally appropriate measures of cognitive function and prevalence of neurocognitive disorder among older Indian women and evaluated the extent to which type of housing (permanent vs. semi-permanent construction materials) modified this effect. Additionally, I estimated the effect of exposure to reported use of unclean cooking fuel within the household on cognitive function and prevalence of neurocognitive disorder among older Indian adults and identified whether this association differed by sex. I used generalized linear models to estimate mean differences in cognitive factor scores and prevalence of neurocognitive disorder by type of cooking fuel within the sample. I added an interaction term between type of cooking fuel and housing type, and type of cooking fuel and sex, to assess for effect modification by type of housing and sex, respectively. Using data from two harmonized cohort studies, Kaiser Healthy Aging and Diverse Life Experiences and Study of Healthy Aging in African Americans, I assessed late-life cognitive function and rate of cognitive change among US male veterans compared with non-veterans, and the extent to which veteran status modified the effect of a self-reported lifetime encounter with blasts/explosions on late-life cognitive function and rate of cognitive change. I used linear mixed effects models with age as the timescale to evaluate the extent to which average cognitive function and rate of cognitive change differed by veteran status. To assess for possible effect modification by veteran status on the association between reported lifetime exposure to blasts/explosions and cognitive measures, I conducted a subgroup analysis by veteran status and a pooled analysis with interaction terms between veteran status, reported lifetime exposure to blasts/explosions and age.

Result: In the population of older Indian women, overall and domain specific cognitive factor scores were worse for women who reported use of unclean cooking fuel compared with women who reported using clean cooking fuel. There was modest evidence that type of housing modified this relationship, namely, women living in homes made with semi-permanent construction materials performed poorer in the specific domains of executive function, language/fluency, and memory (P-value for interaction <0.10). The prevalence ratio (PR) of neurocognitive disorder was modestly consistent with a higher prevalence for women who reported use of unclean cooking fuel (PR: 1.16, 95% CI 0.97, 1.40), but the confidence intervals were consistent with values ranging from little or no effect to moderate increase in prevalence. When assessing whether the association between reported household use of unclean cooking fuel on late-life cognitive function differed by sex, mean differences in cognitive factor scores were slightly greater for women in the overall and specific domains of orientation, language/fluency, and visuospatial ability (P-value for interaction <0.10). Among men, reported household use of unclean cooking fuel was associated with higher prevalence of neurocognitive disorder (PR: 1.41, 95% CI 1.09, 1.83). However, among women, this association was moderately consistent with higher estimates (PR: 1.18, 95% CI 0.93, 1.50), but the confidence intervals were consistent with values ranging from little or no effect to moderate increase in prevalence. Among US men, average verbal episodic memory scores at age 75 years were slightly higher among veterans compared to non-veterans (Coefficient: 0.12, 95% CI 0.01, 0.23). There were no differences in average executive function scores at age 75 years (Coefficient: 0.06, 95% CI -0.06, 0.18) or in the annual rate of change in either cognitive domains comparing veterans to non-veterans. The overall association between reported lifetime exposure to blasts/explosions and late-life cognitive trajectory was null for the full sample and did not reveal robust evidence for effect modification by veteran status.

Conclusion: Among older Indian women, upgrading homes made with semi-permanent materials to permanent materials and transitioning to sustainable use of clean fuel may help reduce risk of late-life cognitive impairment. Further exploration of potential confounders is needed including exploring possible effects of outdoor air pollution and other early and mid-life socio-economic factors to rule out alternative explanations for the association observed among men and to accurately determine differences in effect by sex. In the US, rate of cognitive change by veteran status should remain an important measure to monitor in a rapidly aging veteran population, with further work needed towards correctly assessing potential resilience mechanisms and heterogeneity within the veteran population.