To our knowledge, we are the first group to investigate the seasonality of tuberculosis across the whole of Pakistan and the effect of latitude on this relationship. We report seasonal variation of TB notification across 139 districts in Pakistan, with a peak in TB notifications in quarter two (April-June) and trough in quarter four (October-December). The amplitude of seasonal variation in TB notifications was greater in regions closer to the equator. Seasonal variation in notifications was also more pronounced for extrapulmonary TB than for pulmonary TB.
The results of our study are consistent with those of a smaller study looking at the seasonality of TB in a single district, Lahore, Pakistan, in which the investigators showed a peak in pulmonary TB cases in the second quarter of the year (17). Several studies (6, 9, 12, 22) have explored the relationship between distance from the equator and seasonal variation of TB, and it is plausible that the seasonality of TB is more pronounced in areas where UV exposure is reduced and dermal vitamin D synthesis is low (12, 22). Contrary to a study published in India (12), which showed Northern areas to have greater seasonal variation than those in central southern regions, our study shows reduced seasonal variation in regions further from the equator. Our results are consistent with a study looking at the seasonality of TB in Xinjiang, in which greater seasonality was noted in southern Xinjiang (34ºN-42ºN) than eastern and northern regions (43ºN-48ºN); however, the difference was not statistically significant (9). In the USA, the amplitude of TB seasonality has not been shown to vary according to latitude (6). Wubuli et al. (9) hypothesised that the differences in seasonal variation in TB notifications could be due to the relative latitudes of different countries: all latitudes of continental USA are further from the equator (30ºN-48ºN) and those of Pakistan used in this study (24.5ºN-36.5ºN) are narrower than for the latitudes of India (8ºN-37ºN) and Australia (10ºN-44ºN).
Several hypotheses have put forward to explain the relationship between seasonality and tuberculosis, including winter indoor crowding, the seasonal variation of other respiratory infectious diseases, dietary nutrient intake and dermal synthesis of vitamin D. During the winter season, indoor activities are more common than in a warm climate, which exposes people to tubercle bacilli expelled from infected people in a room with closed windows (26-28). Several other respiratory infectious diseases, both viral and bacterial, peak during the winter (29), which may suppress host immunologic capacity and make people more vulnerable to primary infection or activation of latent TB (6). Seasonal variations in the dietary intake and meal patterns of humans may also affect immune system functions, which may be linked with seasonal variability of TB (26). Vitamin D plays a putative role in the pathogenesis of a variety of respiratory infections (30-32), and it is possible that reduced exposure to sunlight during the winter months (and the consequential reduction in cutaneous vitamin D synthesis) can make a host more susceptible to infection with Mycobacterium tuberculosis (33-35).
Key strengths of this study are that we have pooled outcomes of pulmonary TB and extrapulmonary TB to maximise power. Moreover, we studied the seasonality of pulmonary TB and extrapulmonary TB separately and demonstrated that seasonal variation of extrapulmonary TB was more marked than for pulmonary TB. Data were analysed on a district level, which maximized their granularity. Our analysis also has limitations. Data were not collected on a monthly basis; hence it was not possible for us to identify the specific month/s of the year in which TB peaked or nadired. Data on sex and age were not available for each year, hence we were not able to study whether these variables affected the amplitude of seasonal variation (6) nor if different age groups or sexes had a different peak/trough of TB notifications (10). Other important confounding variables that have been shown to affect the seasonality of TB in other settings include air pollution (37), HIV status (38) and climate parameters, such as rainfall and temperature (26). We were also unable to collect data on the rural or urban nature of the different district. This would be important given a significant body of evidence demonstrates greater UV exposure in rural compared to urban areas (39-41), hence the seasonality of TB might be more pronounced in rural regions. Lastly, we only included smear-positive pulmonary TB patients in this study, hence the results are not applicable to all forms of pulmonary TB.
In conclusion, TB notifications in Pakistan exhibit seasonal variation with a peak in quarter 2 (April-June) and trough in quarter 4 (October-December). The amplitude of seasonality decreases with increasing latitude, and is more pronounced for extrapulmonary TB than for pulmonary TB. The exact cause of the seasonal variation of TB notifications is unknown, but winter indoor crowding, poor UV exposure in winter, seasonal variation in immune function and coinfection with other seasonal pathogens may be responsible. This study may help to shape future public health responses, ensuring that governments can target public health interventions at specific times of the year to reduce TB transmission. Future analyses should ensure confounding variables, such as age, sex, air pollution, temperature and HIV status are controlled for.