Analyzing the seasonality of tuberculosis case notifications in the UK, 2000-2018

Globally, there is seasonal variation in tuberculosis (TB) incidence, yet the biological and behavioural or social factors driving TB seasonality differ across countries. Understanding season-specific risk factors that may be specific to the UK could help shape future decision-making for TB control. We conducted a time-series analysis using data from 152,424 UK TB notifications between 2000 and 2018. Notifications were aggregated by year, month, and socio-demographic covariates, and negative binomial regression models fitted to the aggregate data. For each covariate, we calculated the size of the seasonal effect as the incidence risk ratio (IRR) for the peak versus the trough months within the year and the timing of the peak, whilst accounting for the overall trend. There was strong evidence for seasonality (p < 0.0001) with an IRR of 1.27 (95% CI 1.23-1.30). The peak was estimated to occur at the beginning of May. Significant differences in seasonal amplitude were identified across age groups, ethnicity, site of disease, latitude and, for those born abroad, time since entry to the UK. The smaller amplitude in older adults, and greater amplitude among South Asians and people who recently entered the UK may indicate the role of latent TB reactivation and vitamin D deficiency in driving seasonality.