METHOD: A cross-sectional study was performed in March and April 2016. The outdoor temperatures were measured using the wet-bulb globe temperature (WBGT) tool. The participants completed a self-administered questionnaire containing sociodemographic factors prior to work shift; while working profile, hydration practices, and HRI symptoms at the end of work shift. The hydration status of the respondents was assessed by direct observation of their urine colour. Multiple logistic regression was performed to ascertain the effects of age, working profile, hydration practice, history of previous HRI, and hydration status on the likelihood that outdoor workers having moderate to severe HRI.

RESULTS: A total of 320 respondents completed the questionnaire. The mean (standard deviation) outdoor workplace temperature was 30.5°C (SD 0.53°C). The percentage of respondents who experienced moderate to severe HRI was 44.1%. The likelihood that outdoor workers experienced moderate to severe HRI symptoms was associated with irregular fluid intake [odds ratio (OR): 16.11, 95% confidence interval (95%CI): 4.11; 63.20]; consumption of non-plain water (OR: 5.92, 95%CI: 2.79; 12.56); dehydration (OR: 3.32, 95%CI: 1.92; 5.74); and increasing outdoor workplace temperature (OR: 1.85, 95%CI: 1.09; 3.11).

METHOD: A total of 328 municipal workers were enrolled in April to March 2016 were asked to recall if they experienced eleven HRI symptoms during the previous work day. Rasch Measurement Model was used to examine the unidimensional parameters and bias for gender before identifying the threshold of HRI symptoms. We determined the threshold symptom based on the person-item map distribution on a logit ruler value.

RESULTS: A total of 320 respondents were analysed. The psychometric features HRI symptoms suggested evidence of unidimensionality and free of bias for gender (DIF size =0.57; DIF t value =1.03). Based on the person-item map distribution, the thirst item was determined as the threshold item (Cut-off point = -2.17 logit) for the preventative action purposes to group the person as mild and moderate/severe HRI groups.

METHOD: This study used secondary data retrieved from a cross-sectional study involving 492 male employees' completed data. Eligible participants completed validated questionnaires of the Psychosocial Safety Climate (PSC-12) scale, short version Demand Induced Strain Compensation (DISQ 2.1), Oldenburg Burnout Inventory - Emotional Exhaustion domain and the Three Eating Factor Questionnaire (TEFQ) -Uncontrolled Eating domain; assessing psychosocial safety climate, job demands and job resources, emotional exhaustion, and uncontrolled eating behaviour, respectively. Body mass index (BMI) was calculated based on weight and height. The research statistical model was tested by two-steps of assessment replicating partial least squares structural equation modelling (PLS-SEM).

RESULT: The results show that psychosocial stressors (psychosocial safety climate, job demands and job resources) had significant effects on emotional exhaustion (β= -0.149, p=0.004; β= 0.223, p<0.001; β= -0.127, p=0.013). Emotional exhaustion predicted by work stressors may act as a chain reaction which could result in uncontrolled eating (β=0.138, p=0.005) and high BMI (β=0.185, p<0.001). Emotional exhaustion does mediate the relationship between PSC and uncontrolled eating behaviour (β= -0.021 [95% boot CI bias corrected: -0.048, -0.002]).