We carried out three linked studies to achieve the objectives. (1) We examined data of two large cohort studies. (i) Anhui cohort study (Chen R et al, 2014) we recruited and interviewed 3336 residents aged >60 years from Anhui province, China (1736 aged >65 years in urban) in 2001 and 2003. Using a standard method of the Geriatric Mental State (GMS) interview we recorded disease risk factors and diagnosed dementia. They were followed up to 2011, with 3 health surveys (in 2-4 years interval) to document dementia and causes of death. In 2430 participants aged >65 years without baseline dementia and followed up in the cohort, 271 developed dementia. In a binary logistic regression model, adjusting for age, sex, educational level, lifestyles, body mass index, social network and support and co-morbidities (e.g. stroke, depression) we calculated Odds Ratio (OR) of dementia to estimate the risk of dementia in relation to AP. (a) We compared incident dementia between urban and rural participants as AP in urban China is higher than rural areas, and the analysis was strict to those who had at least primary school education (n=1365) since rural elders had much higher prevalence of being illiterate than urban which affected dementia in relation to AP. 79 urban elders developed dementia (6.3%) and 8 rural (7.2%). Adjusted OR (aOR) analysis showed a 56% increase in dementia risk in urban versus rural, although not statistically significant. (b) We further anlaysed data of the residential distance proximity to major roadways (REDPMR) (an approximate AP exposure) associated with dementia in urban participants. 1455 urban elders without baseline dementia had a mean of REDPMR 280.7 meters (ranged 11-594). Over the 10 years follow-up, 116 developed dementia. The risk of dementia was significantly increased with shorter REDPMR; aOR 1.61 in REDPMR <173 m vs >=265 m. (c) We combined the Anhui data with our earlier meta-analysis (where the Fellow and team systematically searched literature till September 2018 and identified 10 papers - all were published between 2014 and 2018 and from HICs; one case-control study and six cohorts - of which two produced five papers for different air pollutants analysis). Pooled data from eligible studies including the Anhui urban versus rural showed a significantly increased risk of dementia by 26% in the highest vs lowest AP. The Fellow reported its findings in 9th European Public Health and Epidemiology Congress in June 2019 (see photo). (ii) Zhejiang cohort study, 9353 residents aged ≥60 years in Zhejiang province were interviewed to record baseline health and disease risk factors, and measure cognitive function using a standard method in 2014. Daily concentrations of AP in their living areas were obtained from local governments, including PM2.5 PM10, SO2, NO2, CO and O3, and air quality index (AQI) was calculated accordingly. In 2015 and 2016, 8598 survivors were re-interviewed to detect incident cognitive impairment (He F et al, 2020). In 2019, 6115 cohort members were re-examined, with additional diagnosis for Alzheimer’s Disease (AD). 986 AD cases were identified. The AD risk was significantly increased with PM2.5 PM10, CO and AQI (e.g. aOR was 3.92 in PM2.5 >60 vs <=50 μg/m3), but not with SO2, NO2, and O3. Multivariate logistic regression models showed no significant interaction effects of ETS or fish consumption with air pollution on AD; that is, the impact of AP on AD was similar between participants with and with ETS, and also between participants with low and high levels of fish/shrimp consumption (except for PM2.5 exposure, possibly stronger impact in participants with low fish consumption). (2) We carried out a case-control study. Initially we did it in Wolverhampton, UK, having completed all prepared work, but due to the Covid-19 pandemic we seek alternative field. The Fellow completed data collection of 112 cases of dementia and 112 controls in Guangzhou, China using a valid questionnaire we developed in Wolverhampton. The participants with dementia had a significantly shorter REDPMR than their controls (198.6 vs 315.9 meters); aOR of dementia in every meter of REDPMR was 0.992 (see diagram). The analysis did not find any significant interaction effect of ETS or fish consumption with REDPMR on dementia risk. (3) In Guangzhou, the Fellow run two focus groups, using a topic guide based upon the emergent themes from the above quantitative parts and a protocol he developed in Wolverhampton. The Fellow himself presented the focus discussion. The 1st group from six health professionals (age mean 40.8 SD 4.3 years; five men) believed a significant association between AP and dementia although the exact mechanisms of the association is unknown, and considered that reduced exposure of AP in population could be a preventive measure against dementia epidemic. However, in the 2nd group from five family members or caregivers (age 45.8 9.7 years; two men), most of them did not think that AP increased the risk of dementia. They considered that the main causes of dementia would be age and genetic factors, and doing exercise and living with young people could prevent dementia. Nevertheless, both groups thought that air pollution in China was mainly from manufacturing and car emissions, and suggested that to reduce AP it should make more rigorous policies in controlling the numbers of energy-consuming manufactories and private cars. They further believed that planting more trees and reducing vehicle emission would help control AP and prevent dementia.
