Ultra-violet radiation is responsible for the differences in global epidemiology of chickenpox. Part 2
Background
Chickenpox epidemiology is unique among human herpes viruses. In the tropics primary infection is often delayed into later childhood whereas in temperate zones most infection occurs before leaving school. Indeed, in some tropical countries 30-50% of adults are susceptible, compared with only 5-10% from temperate areas.
Conventionally, transmission has been considered to occur by shedding of virus from the upper respiratory tract 1-2 days before the rash. The papers which claim to show such virus transmission however, also conclude that the titres of virus in vesicular fluid are considerably greater than those present in the pharynx and that vesicular virus makes the greatest contribution to spread. Indeed, the few papers cited as providing epidemiological evidence for airborne spread are either mis-quoted, based on case reports or do not reflect the normal transmission environment. In this regard chickenpox appears similar to smallpox, which also had a distinct winter-spring seasonal peak in incidence and was spread partly by the vesicular eruption.
Why such a common, global infection should be less common in children from the tropics when infections are generally more common remains unknown. Although previously suggested factors such as heat, humidity, viral interference, population density or infection with cross-protecting viruses, have been suggested as possible causes of the epidemiological differences, a unified, coherent explanation has eluded discovery. The climatic factor which I propose to show is responsible for the geographical differences in transmission is ultra-violet radiation (UVR). Furthermore, as varicella-zoster virus (VZV) exists only in man, I propose that UVR has been involved in the co-evolution of virus as man migrated out of Africa. The evolution of varying degrees of resistance to UVR among the different genotypes may also have implications for virus reactivation as zoster.
Chickenpox is seasonal in temperate zones, with the highest incidence seen in winter and spring. One explanation for this seasonality could be the significantly higher levels in ultra-violet radiation (UVR) of approximately 10-25-fold seen in summer in temperate zones, which could inactivate virus either in vesicular lesions or after their rupture. Chickenpox is not seasonal in the same way in the tropics possibly because UVR differs only by a factor of two during the year. The tropics however, do experience peaks in chickenpox incidence when the climate is hot, dry and sunny with a rapid decline to very low levels during the rainy season. This appears difficult to reconcile with UVR inactivating virus until the effects of atmospheric pollution on ambient UVR are considered. For example, the Indo-Asian haze, a continent-wide increase in air pollution during the dry season from December to April, has been shown to reduce significantly the level of ambient UVR. As the Monsoon arrives, atmospheric particles and pollutants are washed out, increasing the UVR which inactivates virus more effectively. This correlates very well with the observed chickenpox incidence in Sri Lanka and south India. Furthermore, outbreaks of varicella have been terminated in certain African countries by the arrival of the rainy season. Increased atmospheric pollution might partly explain, in association with locally increased population density, why chickenpox is commoner in urban environments compared with rural communities in adjacent geographic areas.
Further support for the hypothesis derives from sequence analysis which has classified VZV into distinct genotypes. In the largest published study, 348 genotypes of VZV were given geographic locations based on where the virus was originally detected. In the temperate zones which were studied (N America, Argentina, Europe, S Africa, N China, N Asia) a total of 35/259 (13.5%) genotypes were tropical. In contrast, of the 89 isolates from tropical countries/regions (India, Nepal, Bangladesh, Chad, DRC, Southern China, Western Australia, Brazil, Cote D’ Ivoire, Ethiopia, Thailand, Vietnam, Zimbabwe), only 5 (5.6%) were temperate. This difference was statistically significant by Chi-square testing (p < 0.0001). Nevertheless, temperate virus genotypes, which should be more sensitive to UVR than tropical strains, and so would be out-competed in terms of transmission, have been detected in tropical areas, namely Australia, Brazil, Congo, and Mexico City. However, survival of temperate genotypes in these regions is still consistent with the hypothesis when it is considered how reducing ambient UVR allows temperate genotypes to transmit.