Air pollution has been linked to dementia; however, the effects of air pollution on dementia development remain unclear. The novelty of this study was that we examined associations between air pollution and cognitive functions via alterations in sleep cycles. The significance of our results are: (1) air pollution was associated with disruption of sleep quality, (2) air pollution increased the risk of cognitive decline via disrupting the sleep cycle, and (3) changes in brain regions in control of executive, learning, and language functions were associated with air pollution.
This cross-sectional study recruited 4866 subjects who had undergone 1-night PSG measurements in a sleep center in Taipei City. The 1-year average concentrations of air pollution for study subjects in the present study were 24.2 ppb O3, 18.1 ppb NO2, and 29.2 µg/m3 PM2.5 as estimated by the LUR model. Exposure levels of air pollution were similar to those of previous studies conducted in Taipei City. One study estimated levels of O3, NO2, and PM2.5 to be 24.86 ppb, 21.25 ppb, and 22.13 µg/m3, respectively (Zou et al., 2021). Also, a previous report showed that 1-year average concentrations of O3, NO2, and PM2.5 were 27.91 ppb, 18.43 ppb, and 18.44 µg/m3, respectively (Tung et al., 2021c). Levels of O3 and NO2, in this study were relatively lower compared to World Health Organization (WHO) air quality guidelines (Organization, 2016). However, the PM2.5 level was 1.8-fold higher than the 1-year average WHO guideline (10 µg/m3). In this study, subjects were exposed to relatively higher levels of PM2.5; however, effects on their health due to air pollution remain unclear.
We observed that air pollution was associated with disruption of sleep quality, in terms of arousal and the sleep cycle. Similar results were observed in our previous reports, indicating that air pollution exposure (i.e., PM2.5 and NO2) was associated with an increase in arousal (Lo et al., 2021; Tung et al., 2021b). The increased risk of arousal events by air pollution may be associated with structural changes and edema due to inflammation in upper airway tissues (Billings et al., 2019b; Tung et al., 2021a). A previous report suggested that air pollution also causes neuroinflammation which limits the growth or repair of nerve tissues and indirectly interrupts the sleep cycle (Bose et al., 2019). Indeed, we found that air pollution was associated with disruption of the sleep cycle. Alterations in light sleep by an increase in the N1 stage and a decrease in the N2 stage were sensitive to O3 and NO2, whereas PM2.5 was associated with increases in the N2 and N3 stages with decreases in the N1 and deep-sleep REM stages. A previous report showed that PM2.5 was associated with an increase in the N1 stage in OSA patients (Lo et al., 2021). Also, PM2.5 and NO2 were associated with an increase in the N1 stage, whereas a decrease in N2 by NO2 and a decrease in N3 by PM2.5 and NO2 were identified (Tung et al., 2021b). Together, these results suggest that air pollution disrupts the sleep cycle by increasing light sleep during sleep time. Alterations in the sleep cycle due to air pollution could be associated with changes in the upper airway structure. For example, a previous study reported that reduced sleep quality was associated with upper airway edema and inflammation caused by NO2 (Chen et al., 2019). Short-term exposure to O3 may interrupt the sleep cycle since this air pollutant can alter the ratio of ventilation and perfusion in the lungs and result in upper airway obstruction (Weinreich et al., 2015). Furthermore, air pollution regulated the serotonin expression level and damaged gamma-aminobutyric acid (GABA) receptors in the mouse brain (Yokota et al., 2013). The reduction in GABA receptors was associated with an increase in sleep fragmentation of the REM stage (Brooks and Peever, 2011), and decreased serotonin levels shortened the REM duration and increased wake times (Monti, 2011). Taken together, air pollution may cause sleep fragmentation and an increase in light sleep, leading to poor sleep quality.
In the present study, the sleep cycle was associated with cognitive function according to results of the MMSE and CASI. We found that increases in the wake, N1, and N2 stages were associated with cognitive decline, whereas an increase in REM sleep was associated with increases in cognitive functions. These results suggest that sleep quality plays an important role in maintaining cognitive functions. Increases in the wake and N1 percentages were associated with reduced cognitive function by MMSE scores in elderly females (Blackwell et al., 2006). Another study in China suggested that patients with abnormal cognitive function had a higher N1 stage percentage compared to normal participants (Liu et al., 2020). The unique brainwave feature in the N2 stage sleep spindle was associated with memory consolidation, and a decrease in the N2 percentage was associated with cognitive decline (Liu et al., 2020). Furthermore, REM sleep is associated with memory and learning ability (Shatzmiller et al., 2010). One study suggested that sleep with an insufficient REM percentage might not allow processing for memory consolidation, thereby affecting cognitive function (Vertes and Eastman, 2000). Also, a low percentage of REM sleep was associated with an increased risk of developing dementia (Pase et al., 2017). Taken together, the sleep cycle may be a mediator of cognitive function after air pollution exposure.
Next, we investigated the mediating effects of the sleep cycle between air pollution and cognitive function in study subjects. We found that PM2.5 was associated with a decrease in the N1 stage, whereas the N1 stage was associated with a decrease in concentration. Notably, we found the direct effect of PM2.5 was insignificantly associated with concentration. This result suggests that the N1 stage fully mediates between PM2.5 and concentration. Previous studies also showed that PM2.5 was associated with MMSE domains (Aretz et al., 2020; Yao et al., 2021). Also, NO2 and O3 were found to be associated with cognitive function measurements (Chen et al., 2021; Crous-Bou et al., 2020). Therefore, the sleep cycle is essential to maintaining cognitive function and is associated with dementia.
To further understand pathological changes in the brain due to air pollution, we examined associations between air pollution and brain structures. We observed that air pollution was associated with increases in the brain volume and thickness in certain regions associated with executive, learning, and language functions. First, O3 and NO2 were associated with increased volumes of the pars orbitalis, rostral middle frontal, supra marginal, transverse temporal, and pars opercularis. Previous reports indicated that the pars orbitalis is related to language function (Ben Shalom and Poeppel, 2008), the rostral middle frontal is involved in executive function (Nakamura-Palacios et al., 2014), the supra marginal is related to memory function (Rubinstein et al., 2021), and the transverse temporal is involved in language function (Ardila et al., 2016). O3 and NO2 were associated with damage to the integrity of the blood-brain barrier (BBB) (Rivas-Arancibia et al., 2010; Thiel et al., 2001), which elevated oxidative stress and neurotoxicity and disrupted BBB functions (Yan et al., 2015). An in vivo study found that exposure to O3 at 0.25 ppm (4 h/day for 4 weeks) increased the brain volume and caused brain cell and tissue injury (Rivas-Arancibia et al., 2010). Furthermore, we found associations of PM2.5 with an increased volume of the pars triangularis and increased thickness of the fusiform of the brain. The pars trangularis is linked to language function (Yuan et al., 2021), and the fusiform cortical thickness is related to memory function (Zhao et al., 2021). Previous studies showed that PM2.5 was associated with increased thicknesses of the occipital and cingulate lobes (Cho et al., 2020). Also, coarse particulate matter (PM2.5−10) was associated with an increase in the cerebellum volume (Lubczyńska et al., 2020). Brain volume enlargement with oxidative-inflammatory responses was also observed in vivo after 3 and 6 months of exposure to low-level PM2.5 in Taipei City (Shih et al., 2018). Air pollution-induced brain inflammation could cause edema (Shih et al., 2018), leading to sleep cycle disturbances and increased risks of cognitive decline.
There are some limitations to our study. The sample size for cognitive function measurement should be increased in the future. Information on comorbidities was not collected in this study, which should be included in future work. Other cofactors such as socioeconomic status, alcohol, smoking, noise exposure, and physical activities, should be considered in the future. Indoor air pollution and the chemical composition of PM2.5 should be included in future work.