Does skin surface temperature variation account for Buruli ulcer lesion distribution?

Nicola K. Sexton-Oates, Andrew J. Stewardson, Arvind Yerramilli, Paul D.R. Johnson

Research output: Contribution to journalArticleResearchpeer-review

4 Citations (Scopus)


BACKGROUND: Buruli ulcer is a necrotising infection of skin and soft tissue caused by Mycobacterium ulcerans (M. ulcerans). Buruli ulcer most often occurs on limbs, and it is hypothesized this is explained by direct exposure to the environment. However, even on exposed areas Buruli ulcer is not randomly distributed. M. ulcerans prefers an in vitro temperature of 30-33°C and growth is inhibited at higher temperatures. This study investigated whether variations in skin surface temperature distribution in healthy volunteers could partly account for Buruli ulcer lesion distribution. METHODOLOGY/PRINCIPAL FINDINGS: In this observational study, a thermal camera (FLIR E8) was used to measure skin surface temperature at the sternal notch and at 44 predetermined locations on the limbs of 18 human participants. Body locations of high, middle and low Buruli ulcer incidence were identified from existing density maps of lesion distribution. Skin temperature of the three incidence location groups were compared, and differences in age and sex groups were also analysed. We found an inverse relationship between skin temperature and lesion distribution, where high incidence locations were significantly cooler and low incidence locations significantly warmer (Kruskal-Wallis test p<0.0001). Linear mixed effects regression analysis estimated that skin surface temperature accounts for 22.0% of the variance in Buruli ulcer lesion distribution (marginal R-squared = 0.219) in the anterior location group, and 0.6% in the posterior group (marginal R-squared 0.006). Men had warmer upper and lower limbs than females (Mann-Whitney U test p = 0.0003 and p<0.0001 respectively). CONCLUSIONS/SIGNIFICANCE: We have found an inverse relationship between skin temperature and Buruli ulcer lesion distribution, however this association is weak. Additional unknown factors are likely to be involved that explain the majority of the variation in Buruli lesion distribution.

Original languageEnglish
Article numbere0007732
Number of pages14
JournalPLoS Neglected Tropical Diseases
Issue number4
Publication statusPublished - 20 Apr 2020

Cite this