Hodgkin Lymphoma has a seasonal pattern of incidence and mortality that depends on latitude

Abstract

Seasonal variations in incidence and mortality after a Hodgkin lymphoma (HL) diagnosis have been previously described with partly conflicting results. The goal of this analysis is to provide a comprehensive analysis of these seasonal variations. In total, 41,405 HL cases diagnosed between 1973 and 2012 in the 18 Surveillance, Epidemiology, and End Results registries were included. Cosinor analysis and Cox proportional-hazards models were employed to analyze seasonality of incidence and mortality, respectively. HL shows a sinusoid seasonal incidence pattern (p < 0.001). Estimated incidence in March is 15.4% [95%-CI: 10.8-20.0] higher than in September. This sinusoid pattern is more pronounced at higher latitudes (p = 0.023). The risk of dying within the first three years after a HL diagnosis in winter is significantly increased compared to a HL diagnosis in summer at higher latitudes (HR = 1.082 [95%-CI: 1.009-1.161], p = 0.027). Furthermore, increasing northern latitude increases the additional mortality risk conferred by a diagnosis in winter (p_interaction0.033). The seasonality patterns presented here provide epidemiological evidence that Vitamin D might play a protective role in HL. Further evidence on the direct association between Vitamin D levels and the clinical course of HL needs to be collected to advance the understanding of the role of Vitamin D in HL.

Figure 1

Forest plot of the subgroup analysis of seasonality of incidence The seasonality amplitude is shown on the x-axis. Subgroups are shown on the y-axis. The estimated seasonality amplitude with 95% confidence intervals is shown for each subgroup.

Figure 2

Relationship between latitude quartile and estimated seasonality amplitude Latitude quartiles are shown on the x-axis. The first, second, third and fourth latitude quartiles include all counties south of 34.19°N, between 34.19°N and 38.05°N, between 38.05°N and 41.68°N and north of 41.68°N, respectively. The estimated seasonality amplitude of the respective quartile is shown on the y-axis with 95% confidence intervals of the estimated amplitude.

Figure 3

Illustration of the relationship between the peak and trough of vitamin D levels in humans in the northern hemisphere and the overall seasonality pattern of HL incidence The months of the year are shown on the x-axis. The ratio of the respective month’s incidence and the average incidence of all months are shown on the y-axis. The line indicates a fitted curve through the data as estimated by the cosinor model.