Table 2.

Incidence of diabetes in relation to sex-specific quartiles of plasma follistatin levels in MDC-CC.

	Q1	Q2	Q3	Q4	p for trend	HR per SD
Number of participants	1048	1049	1050	1048
Incidence of diabetes n (n per 1000 p-y)	108 (5.16)	132 (6.54)	149 (7.52)	188 (9.87)
Follistatin (NPX)†	17.27 ± 1.21	23.92 ± 1.07	30.06 ± 1.08	42.22 ± 1.23
Model 1 HR	1	1.28 (0.99–1.66)	1.47** (1.15–1.89)	1.97*** (1.55–2.50)	<0.001	1.29*** (1.19–1.40)
Model 2 HR	1	1.12 (0.86–1.45)	1.21 (0.94–1.57)	1.45** (1.13–1.86)	0.003	1.31** (1.11–1.56)
Model 3 HR	1	1.10 (0.85–1.43)	1.16 (0.89–1.50)	1.35* (1.04–1.74)	0.020	1.24* (1.04–1.47)

Model 1: Adjusted for age and sex (n = 4195). Model 2: Adjusted for age, sex, BMI, physical activity, alcohol intake, fiber intake, smoking habits, use of antihypertensive medications, systolic blood pressure, LDL, HDL cholesterol, use of lipid lowering medications, fasting glucose (n = 4060). Model 3: Model 2 and CRP (n = 4060). The relationship between follistatin levels in plasma and incidence of diabetes during the follow up was explored in 4195 individuals, and fully-adjusted association study was performed in 4060 individuals. HR, hazard ratio. *p < 0.05, **p < 0.01, ***p < 0.001. ^†Follistatin is expressed as linear Normalized Protein eXpression (NPX) AU for relative quantification according to Olink guidance.