Recent pesticide exposure affects sleep: A cross-sectional study among smallholder farmers in Uganda

BACKGROUND: Poorly educated smallholder farmers in low-income countries are highly exposed to pesticides. This can result in adverse mental health issues, of which sleep problems might be an underlying indicator. We aim to examine the association between sleep problems and pesticide exposure among smallholder farmers in Uganda.

METHODS: A cross-sectional survey with 253 smallholder farmers was conducted between October and December 2019. Sleep problems were assessed during the week before the visit using the Medical Outcomes Study Sleep Scale (MOS-SS). Exposure to pesticides was assessed as application days of any pesticide and as use of 2,4-D, glyphosate, mancozeb, organophosphates & carbamates, pyrethroids and other pesticides during the week and year prior to the visit. Associations were assessed using adjusted multivariable logistic regression models.

RESULTS: Increased odds ratio (OR) for the sleep problem index 6-items (OR [95% Confidence Interval] 1.99 [1.04; 3.84] and 3.21 [1.33; 7.82]), sleep inadequacy (1.94 [1.04; 3.66] and 2.49 [1.05-6.22]) and snoring (3.17 [1.12; 9.41] and 4.07 [1.04; 15.14]) were observed for farmers who respectively applied pesticides up to two days and three or more days in the past week compared to farmers who did not apply during the past week. Gender-stratified analyses showed a higher OR for female applicators (4.27 [1.76-11.16]) than for male applicators (1.82 [0.91-3.79]) for the association between the sleep problem index 6-items and pesticide use in the week before the visit. Increased ORs were also observed for the association between the sleep problem index 6-item and mancozeb exposure during the past year 2.28 [1.12-4.71] and past week 2.51 [0.86-7.55] and glyphosate exposure during the past week 3.75 [1.24-11.8] compared to non-applicators.

DISCUSSION: Our findings suggest an increased risk of sleep problems among smallholder farmers in a pesticide-exposure-dependent way in a low-income context. Further gender-stratified, longitudinal investigations are warranted to confirm these findings.