To investigate the relationship between serum 25-hydroxyvitamin D (25(OH)D) and risk of incident diabetes in Hong Kong Chinese, after accounting for the effect of multiple bone- and mineral-related markers.

We conducted a retrospective study on the Hong Kong Osteoporosis Study cohort. Incident diabetes was ascertained using electronic medical records. Serum 25(OH)D was measured at baseline and its association with incident diabetes was evaluated using multivariable Cox proportional-hazard regression.

Individuals (n 4342) aged 20 years or above (1395 men, 2947 women; mean age 54·3 (sd 16·5) years) from the Hong Kong Osteoporosis Study, who were free of diabetes at baseline, were included.

During 40 124·7 person-years of follow-up (a median of 9·2 years), 443 participants developed diabetes. Mean 25(OH)D was 63·34 (sd 13·07) nmol/l. Age-, sex- and BMI-adjusted Cox proportional-hazard regression showed no significant difference in the risk of incident diabetes between the lowest and the highest quintiles of 25(OH)D. In the analysis of the interaction effect between 25(OH)D and serum Ca, the interaction term did not affect the risk of incident diabetes significantly (P = 0·694). Similarly, there was no significant interaction of different subgroups (age, sex, BMI, femoral-neck T-score, serum Ca levels) with serum 25(OH)D.

The present study finds that serum vitamin D level is not associated with the risk of incident diabetes in Hong Kong Chinese and this relationship is not modified by serum Ca level.