Overall sleep status and high sensitivity C-reactive protein: a prospective study in Japanese factory workers

We investigated the relation between overall sleep status based on the modified Pittsburgh Sleep Quality Index (PSQI) global score and subsequent changes in serum high-sensitivity C-reactive protein (hsCRP) in a population of Japanese factory workers, who were predominantly female. A total of 991 Japanese with inflammation classified as low cardiovascular risk (baseline hsCRP < 1.0 mg L(-1) ) were grouped according to the presence or absence of unfavourable sleep, defined as a modified PSQI global score > 5.5 points. The subsequent changes in hsCRP after 3 years were then compared in the two groups. Analysis of covariance incorporating log-transformed baseline hsCRP, age, sex, lifestyle and physical and biochemical profiles was used to compare the geometric means of hsCRP at year 3 in each sleep status group. A logistic regression model incorporating the same variables was used to calculate the odds ratios for development of inflammation with a medium-to-high cardiovascular risk (hsCRP at year 3 ≥ 1.0 mg L(-1) ) comparing the presence or absence of unfavourable sleep habits. The multivariate-adjusted geometric mean of hsCRP at year 3 was significantly higher in subjects with unfavourable sleep habits compared with those with a normal pattern (0.275 versus 0.242 mg L(-1) ). The multivariate-adjusted odds ratio for developing increased and potentially pathogenic levels of inflammation due to unfavourable sleep was 2.08 (95% confidence interval = 1.29-3.35). There was a significant linear trend for the development of increased inflammation across the modified PSQI global scores (P = 0.04). Unfavourable sleep is associated with activation of low-grade systemic inflammation.