Increased incidence of Lyme borreliosis following mild winters and during warm, humid summers

It was with great interest that we read the article by Bennet et al. [1], in which they analysed the influence of climatic features on the incidence of Lyme borreliosis during 1997– 2003 in a defined area of southern Sweden. We carried out a similar analysis in Scotland for the years 1999–2003 and published our results in 2005 [2]. The remarkable finding is that although the two study designs were significantly different, but in overlapping years, similar conclusions were drawn. Bennet et al. [1] derived the incidence of Lyme borreliosis from the cases of erythema migrans recorded in the patients’ medical records, whereas we based it on laboratory-confirmed cases. As erythema migrans (EM) has a relatively close temporal relationship with the onset of infection and tick bite, Bennet et al. [1] analysed their incidence rates at monthly intervals and expressed them as cases per 100,000 population; in this way they effectively analysed the whole study population. In our study, the temporal relationship was less tightly defined, since only a portion of our cases demonstrated early Lyme disease with EM. Also, because we were aware that the rates of samples referred to our laboratory were not comparable in different years and from different parts of the country, we measured infection rates as cases/100 samples. We divided the country into three areas with distinctly different climatic conditions and compared the infection rates. As in the Swedish study, we were able to identify that colder winters depressed the incidence of Lyme disease in the succeeding summer. This effect was most evident in 2001, in the northern and eastern regions of Scotland. The effect of excess summer rainfall was also noted. The western region, which usually has wetter summers, especially compared to the east, had a significantly lower incidence than the other two regions in all years studied. However, in 2000, a significantly drier summer coincided with a significantly higher infection rate. Like Bennet et al. [1], we suggested this peak could be attributed to increased tick exposure as a result of increased outdoor activity. However, in our country we had an additional factor supporting the contribution of human activity, which would not apply to the Swedish study. In 2001, a major foot-andmouth disease epidemic led to countryside access restrictions and depressed visitor numbers, particularly in the eastern part of the country in June and July, as reported in the Scottish Visitor Attraction Barometer for 2000 and 2001 (http://www.scotexchange.net). The resulting reduction in tick exposure, combined with the effects of the preceding cold winter on the tick population, may explain the reduced infection rate in this region. Identifying the factors that most influence the incidence of Lyme borreliosis is very difficult, so the additional evidence provided by the Swedish study in support of rational explanations is welcomed.