Background At present, domestic and foreign studies on the association between dietary magnesium and diabetes risk are not consistent, and there are relatively few prospective studies in China and the study population is relatively limited.

Objective To explore the association between dietary magnesium intake and diabetes risk in Chinese adults in 15 provinces (autonomous regions, municipalities), and to provide a scientific basis for revising dietary magnesium intake reference for Chinese residents.

Methods A total of 8061 adults aged 18-64 who participated in at least two follow-up surveys in the China Health and Nutrition Survey in 2009, 2015, and 2018, had complete survey data, and did not report diabetes at baseline were selected as subjects. Food consumption data were collected from 3-day 24-hour dietary recalls and by weighing household cooking oil and condiments. The average daily dietary magnesium intake was calculated based on the food composition table. Multiple Cox proportional risk regression model and restricted cubic spline (RCS) model were used to analyze the association and dose-response relationship between dietary magnesium intake and diabetes risk. Diabetes was defined according to the Chinese Guidelines for the Prevention and Treatment of Type 2 Diabetes (2020 edition).

Results A total of 47237.46 person-years were followed up, with an average follow-up of 5.86 years. Among 8061 subjects, the incidence rate was 8.86%. Compared with those in the top quintile of magnesium intake (Q5), those with lower dietary magnesium intake were more likely to be female, have higher income, higher education, live in urban areas, and have lower intakes of energy, dietary fiber, and dietary calcium. After adjusting for demographic characteristics, lifestyle, and dietary factors, compared with adults in the lowest quintile of dietary magnesium intake, the results of Cox proportional risk regression model showed that the second (median: 220.96 mg·d⁻¹), third (median: 263.01 mg·d⁻¹), and fourth (median: 312.33mg·d⁻¹) quintile dietary magnesium intake reduced the risk of diabetes by 45% (HR=0.55, 95%CI: 0.43-0.71), 39% (HR=0.61, 95%CI: 0.47-0.78), and 34% (HR=0.66, 95%CI: 0.51-0.78), respectively. The results of RCS analysis showed that dietary magnesium intake and the risk of diabetes were U-shaped overall. Taking the 5th percentile magnesium intake as reference, when dietary magnesium intake was lower than 240 mg·d⁻¹, the risk of diabetes gradually decreased with the increase of magnesium intake; the risk was the lowest at 240 mg·d⁻¹, followed by a slight increase in risk at 240-400 mg·d⁻¹; and no statistical difference presented in the association between dietary magnesium and diabetes risk after 650 mg·d⁻¹.

Conclusion The study findings suggest an association between dietary magnesium intake and diebetes risk. The association is negative and non-linear when dietary magnesium intake is below 240 mg·d⁻¹.