Assessing the exposure-response relationship of sleep disturbance and vibration in field and laboratory settings.

Exposure to nocturnal freight train vibrations may impact sleep, but exposure-response relationships are lacking. The European project CargoVibes evaluated sleep disturbance both in the field and in the laboratory and provides unique data, as measures of response and exposure metrics are comparable. This paper therefore provides data on exposure-response relationships of vibration and sleep disturbance and compares the relationships evaluated in the laboratory and the field. Two field studies (one in Poland and one in the Netherlands) with 233 valid respondents in total, and three laboratory studies in Sweden with a total of 59 subjects over 350 person-nights were performed. The odds ratios (OR) of sleep disturbance were analyzed in relation to nighttime vibration exposure by ordinal logit regression, adjusting for moderating factors common for the studies. Outcome specific fractions were calculated for eleven sleep outcomes and supported comparability between the field and laboratory settings. Vibration exposure was significantly associated with sleep disturbance, OR = 3.51 (95% confidence interval 2.6-4.73) denoting a three and a half times increase in the odds of sleep disturbance with one unit increased 8 h nighttime log10 Root Mean Square vibration. The results suggest no significant difference between field and laboratory settings OR = 1.37 (0.59-3.19). However, odds of sleep disturbance were higher in the Netherlands as compared to Sweden, indicating unexplained differences between study populations or countries, possibly related to cultural and contextual differences and uncertainties in exposure assessments. Future studies should be carefully designed to record explanatory factors in the field and enhance ecological validity in the laboratory. Nevertheless, the presented combined data set provides a first set of exposure response relationships for vibration-induced sleep disturbance, which are useful when considering public health outcomes among exposed populations.