Fig. 1. Opportunities for and risks of accidental self-laceration among Tsimane by age and sex.

Opportunities for laceration are estimated using time allocation data. Stacked bars show proportion of time at risk of self-laceration (n = 55,035 observations of 904 individuals across six villages) from either food processing (e.g., butchering), other household tasks (e.g., sharpening knives or clearing trails with machetes), labor (e.g., building a house or boat, tree chopping, or clearing brush in fields), tool manufacture including children’s play with machetes, and resource acquisition (e.g., harvesting with knives or machetes and hook-and-line fishing). Activities for which there was uncertainty over whether one faced self-laceration risk were omitted from the risk set. Risks of laceration are estimated using a hazard function based on retrospective self-reports among a different Tsimane sample collected contemporaneously as the time allocation data (n = 388 individuals aged 11 to 75 years across 16 villages). The hazard function is modeled separately for each sex using the bshazard package in R, stratifying by episode number to account for repeated lacerations within individuals over time. The blue line shows the mean hazard for each 5-year age interval. F, female; M, male.