Vagal tone mediates the association between work stress and glycemic status: results from the Mannheim Industrial Cohort Study (MICS)

Introduction: The WHO has declared work related stress as one of the primary challenges in the 21st century. More important, work stress is recognized as a significant determinant for health. We previously showed that work stress is associated with an increased risk of pre-diabetes and diabetes in apparently healthy workers. Likewise, an increased perception of work stress is negatively associated with autonomic nervous system (ANS) function as indicated by heart rate variability (HRV). We investigate to what extend the association between work stress and glycemic status is mediated through HRV. Methods: We utilized a large secondary data set from the Mannheim Industrial Cohort Study, conducted by Health Vision Ltd.. A total of 13 study sites (Companies from the secondary sectors) from all over Germany with a total of 9924 participants collected between 2010 – 2012 were available. The root mean squared successive differences (RMSSD) were averaged from long-term heart rate monitoring (beat to beat) for work time and nighttime as indicators of vagal tone. Due to skewed data HRV measures were log-transformed. Glycemic status was indexed by fasting plasma glucose (FPG) and glycosylated hemoglobin (HbA1c). Work stress was indexed by Effort-Reward Imbalance (ERI) to quantify the degree of failed reciprocity of social interaction in the work context. Stress is apparent if an imbalance between high ''cost'‘ and low ''gain'‘ exists (ratio > 1). We used structural equation modeling (SEM), with maximum likelihood with missing values as well as the clustered sandwich estimator relaxing the assumption of independence of observations between study sites. Through all models ERI served as independent, RMSSD as mediator and FPG or HbA1c as dependent variable. All models were additionally adjusted for age and sex. Results: The final sample consists of 9924 individuals (mean age 41.9 ± 10.9, 19% females). In all four models, ERI is significantly negatively associated with log(RMSSD) (path A) at day and nighttime for both FPG and HbA1c. ERI is significantly positively associated with HbA1c if log(RMSSD) is not in the model (path C). The path C‘ is no longer significant when log(RMSSD) is introduced as mediator in the FPG but not in the HbA1c models. Log(RMSSD) at day and nighttime is significantly negatively associated with FPG and HbA1c. SEM results indicate that 9.5% (HbA1c) or 24% (FPG) of the total effect appears to be mediated in the daytime models and in the nighttime models 6.7% (HbA1c) or 14.1% (FPG). Discussion: In our study we investigate to what extend the association between work stress as indicated by the Effort Reward Imbalance and glycemic status is mediated through vagal tone as indicated by RMSSD. Our results suggest that glycemic status is partly mediated through vagally mediated HRV. On the basis of this data we conclude that work stress may be an additional factor promoting a diabetic metabolic state. If supported by prospective evidence, results point to a new approach towards primary prevention of diabetes.