Objectives: Illness can disrupt training and competition performance of athletes. Few studies have quantified the relative contribution of the known medical, behavioural and lifestyle risk factors.

Design: Cross-sectional.

Methods: Olympic athletes from 11 sports (n=221) were invited to complete questionnaires administered nine months before the Rio 2016 Olympic Games. These included the Depression, Anxiety and Stress Questionnaire (DASS-21), Perceived Stress Scale (PSS), Dispositional Resilience Scale (DRS), Recovery-Stress Questionnaire (REST-Q-52 item), Low Energy in Females Questionnaire (LEAF-Q), a modified Personal and Household Hygiene questionnaire, Epworth Sleepiness Scale, Pittsburgh Sleep Quality Index, and custom-made questionnaires on probiotic usage and travel. An illness (case) was defined as an event which limited training or competition for greater hours in the prior month. Odds ratios and attributable fractions in the population (AFP) were utilised for categorical variables with independent t-tests or Wilcoxon rank-sum for continuous variables.

Results: Eighty-one athletes responded (male, n=26; female, n=55). There were 16 illness cases and 65 controls. Female athletes were at higher odds of illness (OR=9.4, 95%CI 1.3-410, p=0.01, AFP=0.84). Low energy availability (LEAF-Q score ≥8: OR=7.4, 95%CI 0.78-352, p=0.04, AFP=0.76), depression symptoms (DASS-21: depression score >4, OR=8.4, 95%CI 1.1-59, p<0.01; AFP=0.39) and higher perceived stress (PSS: 10-item, p=0.04) were significantly associated with illness.

Conclusions: Female sex, low energy availability, and mental health are associated with sports incapacity (time loss) due to illness. Low energy availability had high attributable fractions in the population and stands out as a primary association with illness.