Global burden of non-melanoma skin cancers among older adults: a comprehensive analysis using machine learning...

Non-melanoma skin cancers (NMSCs), including basal cell carcinoma (BCC) and squamous cell carcinoma (SCC), have shown significant global increases in burden, particularly among older adults, with wide regional, gender, and socio-demographic disparities. However, no previous study has comprehensively analyzed these trends using integrated multi-model approaches. A comprehensive analysis of global non-melanoma skin cancer (NMSC) data from 1990 to 2021 was performed, including trend analysis, frontier analysis, and decomposition analysis to compare the burden of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) across regions, genders, and socio-demographic groups. Temporal trends in incidence, prevalence, and disability-adjusted life years (DALYs) were identified using epidemiological decomposition techniques. To forecast future disease burden from 2022 to 2050, eight machine learning models (Prophet, ARIMA, TBATS, ElasticNet, ETS, VAR, HoltWinters, and Theta) were applied, and the most optimal predictive models for BCC and SCC were selected based on performance metrics. From 1990 to 2021, the global burden of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) among older adults increased substantially. The incidence of BCC rose by 291.85%, with an age-standardized rate (ASR) increase of 62.79%, while SCC incidence increased by 322.77%, with a 67.42% rise in ASR. High Socio-Demographic Index (SDI) regions showed the largest ASR increases for BCC (102.73%) and SCC (107.31%), whereas low SDI regions demonstrated more modest changes or declines (BCC: - 2.86%, SCC: - 3.08%). Gender disparities were observed, with males accounting for 56.47% of SCC cases globally. Decomposition analysis revealed that population growth was the primary driver of increased NMSC burden, contributing 62.52-83.92% of the rise in incidence and DALYs, while epidemiological changes accounted for up to 52.2% in high SDI regions. Frontier analysis highlighted significant regional disparities, with high SDI regions, such as North America, exhibiting the highest age-standardized incidence rates for BCC (2,923.22 per 100,000) and SCC (1,805.33 per 100,000), whereas low SDI regions, like sub-Saharan Africa, showed minimal burdens near frontier values. Finally, disease burden forecasting for 2022 to 2050 using machine learning models predicted a slight decrease in the incidence rates of BCC (from 314.00 to 308.73 per 100,000) and SCC (from 157.67 to 148.16 per 100,000). Disability-adjusted life years (DALYs) are projected to decline for SCC (79.32 to 75.20 per 100,000) but increase slightly for BCC (0.136 to 0.151 per 100,000). These projections underscore the need for tailored strategies to address the growing burden of NMSC. The growing global burden of NMSCs, particularly in high SDI regions and among older adults, underscores the urgent need for tailored prevention, resource allocation, and management strategies. Machine learning-based approaches provide robust insights into global trends and future challenges for BCC and SCC.