Using innovative social media insights and satellite imagery, researchers uncover how poorly planned nighttime lighting disrupts sleep and poses a growing public health concern in urban and rural China.
Study: Outdoor Artificial Light at Night and Insomnia-Related Social Media Posts. Image Credit: YIUCHEUNG / Shutterstock
In a recent study published in the JAMA Network Open, researchers investigated the association between outdoor artificial light at night (ALAN) exposure and insomnia incidence using social media data and satellite-derived nighttime light images in mainland China.
Background
Light pollution, driven by poorly designed artificial lighting, is rapidly increasing, with China experiencing over 6% annual growth in nighttime light. ALAN disrupts circadian rhythms, inhibits melatonin production, and diminishes sleep quality, potentially contributing to depression, insomnia, and metabolic disorders.
Traditional methods of studying ALAN’s impact on insomnia, such as surveys, are prone to recall bias and limited coverage. By integrating social media data and satellite-based ALAN measurements, this study introduces an innovative and real-time approach to understanding ALAN’s health impacts. Further research is essential to refine these methodologies and address broader environmental and health implications.
About the Study
In the present study, following Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines, insomnia prevalence was estimated through daily insomnia-related posts on Weibo, a major Chinese social media platform with 257 million daily active users. A two-stage data collection process employed the Scrapy framework, leveraging internet protocol location functionality introduced by Weibo in 2022. Posts containing insomnia-related keywords were identified and supplemented by user demographic data. Extreme gradient boosting (XGBoost) was used to classify relevant posts, addressing potential misclassification from unrelated content like advertisements.
The analysis period spanned May 2022 to April 2023, with city-level data aggregated daily. To ensure comparability, insomnia incidence was calculated as the number of insomnia-related posts per 10,000 users aged 15-39, constituting 96% of the platform’s user base. ALAN exposure, quantified through National Aeronautics and Space Administration (NASA)’s Black Marble nighttime light remote sensing images, provided daily intensity estimates at a 500-meter resolution. This high-resolution dataset allowed the researchers to analyze ALAN exposure at a granular level, uncovering subtle geographical and temporal trends. Missing data were imputed using temporal averages, and extreme values were trimmed to ensure reliability.
Meteorological variables, air quality indices, and social media trends were incorporated as covariates. Statistical methods included Pearson correlation, multivariable regression, and smoothing spline models to evaluate exposure-response relationships. Subgroup and sensitivity analyses confirmed the accuracy of the results. Although these methods address some biases, the exclusion of indoor ALAN exposure, screen usage, and non-public social media posts highlights the need for complementary data sources in future studies.
Study Results
The study analyzed 1,633,151 social media posts from 336 cities in mainland China, ultimately including 1,147,583 posts identified as insomnia-related after thorough data processing. During the study period from May 2022 to April 2023, daily mean ALAN exposure ranged from 3.1 to 221.0 nW/cm²/sr. Spatially, ALAN exposure exhibited a distinct pattern, with higher intensities observed in eastern regions and urban centers, mainly provincial capitals and economically developed cities. This geographic distribution mirrored the incidence of insomnia, as reflected in the frequency of insomnia-related posts. A statistically significant correlation between ALAN exposure and insomnia-related posts was evident across all subgroups.
Regression analysis revealed an association between ALAN exposure and insomnia incidence. In the unadjusted model, every 5 nW/cm²/sr increase in ALAN exposure corresponded to a 0.390% increase in insomnia incidence. After adjusting for multiple covariates, this association remained significant, with a 0.377% increase per 5 nW/cm²/sr. Stratifying cities by ALAN exposure quartiles showed a progressive rise in insomnia incidence, ranging from 0.569% in the second quartile to 4.320% in the highest quartile compared to the reference group.
The exposure-response analysis highlighted a nonlinear relationship, with the steepest increases in insomnia incidence at lower ALAN exposure levels, plateauing at higher levels. Interestingly, nearly two-thirds of the cities had ALAN exposure levels between 10 and 80 nW/cm²/sr, falling within the range of rapid incidence growth. This finding underscores the urgency for targeted interventions in these regions.
Subgroup analysis demonstrated variability in the association between ALAN and insomnia. Medium and small cities exhibited higher susceptibility to ALAN exposure, with increases of 0.603% and 0.622% in insomnia incidence per 5 nW/cm²/sr, respectively, compared to 0.284% in larger cities. The absence of robust lighting policies in less developed cities likely exacerbates this disparity, emphasizing the need for urban planning reforms. Seasonal variations, extreme temperature conditions, and periods of poor air quality intensified the observed effects.
Sensitivity analyses confirmed the accuracy of these findings. Adjusting for factors like per capita Gross Domestic Product (GDP) or social media popularity yielded consistent results. No false-positive associations were detected in posts unrelated to insomnia.
Conclusions
To summarize, this ecological study utilized large-scale social media data from 336 cities in China, examining 1,147,583 insomnia-related posts from May 2022 to April 2023. The findings demonstrate a significant association between increased ALAN exposure and higher insomnia incidence, particularly in small-to-medium cities.
Proposed mechanisms include melatonin suppression, circadian rhythm disruption, and oxidative stress activation. Unlike in developed countries, disparities in urban development and poorly planned lighting contribute to these risks in China. Policymakers should consider implementing localized light pollution regulations and adopting sustainable lighting designs to mitigate these health risks while fostering equitable urban development.
Future studies should address this research’s limitations, including the role of indoor lighting and other potential confounding factors, to provide a more comprehensive understanding of ALAN’s impact on public health.
