H Korthals Altes1, K Kremer, C Erkens, D van Soolingen, J Wallinga. 1. Epidemiology and Surveillance, Tuberculosis and M/XDR-TB Programme, National Institute for Public Health and the Environment, Bilthoven, The Netherlands. hester.korthals.altes@rivm.nl
Abstract
SETTING: The seasonality of tuberculosis (TB) incidence suggests that the risk of infection or development of disease has a seasonal component. OBJECTIVE: To investigate factors associated with seasonal patterns of TB disease in the Netherlands by splitting notifications according to origin (natives vs. non-natives) and disease site (pulmonary TB [PTB] vs. extra-pulmonary TB [EPTB]). We focus on the presence of a seasonal peak, as much debate has centred on factors enhancing transmission vs. disease development. DESIGN: Monthly notifications were derived from culture sample dates of all cases between 1993 and 2008. We fitted seasonal autoregressive integrated moving average (SARIMA) models to the time series. Seasonal decomposition revealed seasonal trends. To assess the seasonality of the peak, we repeated the analysis omitting December (trough) notifications. RESULTS: TB notifications show a seasonal pattern, with a peak in spring and a trough in winter, which is present in both PTB and EPTB and in both natives and non-natives. However, when excluding December notifications, seasonality only holds in non-native EPTB and non-native TB notifications. CONCLUSION: A seasonal peak in TB notifications (March-June) is apparent in non-natives, but is absent in natives. This peak is driven by the seasonality of EPTB notifications, which are highest in June-July. The contribution of winter crowding is discussed. Vitamin D deficiency, enhancing disease development at the end of winter-early spring, seems the most likely factor explaining the yearly peak in EPTB.
SETTING: The seasonality of tuberculosis (TB) incidence suggests that the risk of infection or development of disease has a seasonal component. OBJECTIVE: To investigate factors associated with seasonal patterns of TB disease in the Netherlands by splitting notifications according to origin (natives vs. non-natives) and disease site (pulmonary TB [PTB] vs. extra-pulmonary TB [EPTB]). We focus on the presence of a seasonal peak, as much debate has centred on factors enhancing transmission vs. disease development. DESIGN: Monthly notifications were derived from culture sample dates of all cases between 1993 and 2008. We fitted seasonal autoregressive integrated moving average (SARIMA) models to the time series. Seasonal decomposition revealed seasonal trends. To assess the seasonality of the peak, we repeated the analysis omitting December (trough) notifications. RESULTS: TB notifications show a seasonal pattern, with a peak in spring and a trough in winter, which is present in both PTB and EPTB and in both natives and non-natives. However, when excluding December notifications, seasonality only holds in non-native EPTB and non-native TB notifications. CONCLUSION: A seasonal peak in TB notifications (March-June) is apparent in non-natives, but is absent in natives. This peak is driven by the seasonality of EPTB notifications, which are highest in June-July. The contribution of winter crowding is discussed. Vitamin D deficiency, enhancing disease development at the end of winter-early spring, seems the most likely factor explaining the yearly peak in EPTB.
Authors: J Jubulis; A Kinikar; M Ithape; M Khandave; S Dixit; S Hotalkar; V Kulkarni; V Mave; N Gupte; A Kagal; S Jain; R Bharadwaj; A Gupta Journal: Int J Tuberc Lung Dis Date: 2014-02 Impact factor: 2.373