Abstract
PURPOSE: Large-scale datasets of government workers have been increasingly used to generate robust epidemiological evidence. However, it remains unclear whether this specific occupational group is representative of the general population in health-related data. This study aimed to examine whether health checkup data from government workers are representative of the general population. METHODS: We analyzed health checkup data from the Mutual Aid Association (MAA) for local government employees and from the National Database of Japan (NDB) across 46 prefectures. Sex-specific and age-adjusted prevalences (AAPs) of health examination abnormalities, medication use, and health behaviors were calculated for each prefecture. Correlation coefficients between the two populations were then determined to evaluate the similarity of regional patterns. RESULTS: Among men, all health examination abnormalities were significantly less prevalent in MAA than in NDB. Among women, the prevalence was also lower in MAA than in NDB, except for HbA1c and liver dysfunction. The MAA population showed significantly lower medication use for hypertension and diabetes mellitus, but higher use for dyslipidemia. Unfavorable health behaviors such as smoking, weight gain, and habitual breakfast skipping were less common in MAA, whereas favorable behaviors such as regular exercise and physical activity were more frequent among men. Correlation coefficients of AAPs ranged from 0.36 to 0.88 for examination abnormalities, 0.21 to 0.80 for medication use, and 0.35 to 0.73 for health behaviors. CONCLUSION: Significant positive correlations were observed between MAA and NDB in the prevalence of health examination abnormalities and health behaviors, indicating that regional health patterns among government workers closely mirror those of the general population. These findings suggest that health data from government workers can serve as a valuable resource for assessing regional differences in population health in Japan.