Table 4.

Analysis of effect modification: adjusted hazard ratios^a and 95% confidence intervals (CI) for the development of type 2 diabetes according to quartiles of predicted serum vitamin D in the SUN cohort stratified by potential confounders (age, sex, and obesity)

	N	Incident T2D	Q1	Q2	Q3	Q4	p for interaction
Age							0.120
Age ≥ 50 years	3292	127	1.00 (reference)	0.63 (0.38–1.03)	0.78 (0.44–1.36)	0.70 (0.34–1.43)
Age < 50 years	15,302	82	1.00 (reference)	0.73 (0.40–1.34)	0.71 (0.34–1.50)	0.23 (0.06–0.85)
Sex							0.961
Women	11,256	49	1.00 (reference)	0.78 (0.33–1.85)	0.96 (0.35–2.62)	0.54 (0.13–2.20)
Men	7338	160	1.00 (reference)	0.57 (0.37–0.89)	0.63 (0.38–1.05)	0.45 (0.23–0.90)
Obesity							0.806
Obese	844	73	1.00 (reference)	0.77 (0.31–1.93)	1.69 (0.59–4.85)	0.94 (0.11–8.22)
Non-obese	17,750	136	1.00 (reference)	0.67 (0.43–1.03)	0.65 (0.39–1.08)	0.49 (0.25–0.95)

^aAdjusted for sex, age, marital status, smoking status (current, former or never smoker), smoking pack-years, weight change, years of university, TV hours/day, family history of T2D, physical activity (MET-h/week), Trichopoulou’s 9-point score/Mediterranean dietary pattern, energy intake (kcal/day), sugar-sweetened beverage consumption (servings/day), snacking, following a special diet, prevalent hypertension, prevalent cancer, prevalent hypercholesterolemia, prevalent hypertriglyceridemia, obesity (kg/m²)