Glycerol and fatty acids in serum predict the development of hyperglycemia and type 2 diabetes in Finnish men

Abstract

Objective: We investigated the association of fasting serum glycerol and fatty acids (FAs) as predictors for worsening of hyperglycemia and incident type 2 diabetes.

Research design and methods: Cross-sectional and longitudinal analyses of the population-based METabolic Syndrome in Men (METSIM) Study included 9,398 Finnish men (mean age 57 ± 7 years). At baseline, levels of serum glycerol, free FAs (FFAs), and serum FA profile, relative to total FAs, were measured with proton nuclear magnetic resonance spectroscopy.

Results: At baseline, levels of glycerol, FFAs, monounsaturated FAs, saturated FAs, and monounsaturated n-7 and -9 FAs, relative to total FAs, were increased in categories of fasting and 2-h hyperglycemia, whereas the levels of n-3 and n-6 FAs, relative to total FAs, decreased (N = 9,398). Among 4,335 men with 4.5-year follow-up data available, 276 developed type 2 diabetes. Elevated levels of glycerol, FFAs, monounsaturated FAs, and saturated and monounsaturated n-7 and -9 FAs, relative to total FAs, predicted worsening of hyperglycemia and development of incident type 2 diabetes after adjustment for confounding factors. n-6 FAs, mainly linoleic acid (LA), relative to total FAs, were associated with reduced risk for the worsening of hyperglycemia and conversion to type 2 diabetes.

Conclusions: Our large population-based study shows that fasting serum levels of glycerol, FFAs, monounsaturated FAs, saturated FAs, and n-7 and -9 FAs are biomarkers for an increased risk of development of hyperglycemia and type 2 diabetes, whereas high levels of serum n-6 FAs, reflecting dietary intake of LA, were associated with reduced risk for hyperglycemia and type 2 diabetes.

Figures

Figure 1

Mean values and their 95% CIs of fasting levels of glycerol (A and B) and FFAs (C and D) across the entire range of fasting and 2-h glucose categories in the cross-sectional METSIM Study cohort. P values (from ANOVA post hoc tests) indicating statistical significance with respect to the reference category (fasting or 2-h plasma glucose <5.0 mmol/L) are coded as follows: **P < 0.01. P values for trends, adjusted for age and BMI, were as follows: 4.5 × 10−28(A), 1.2 × 10−99(B), 4.3 × 10−51(C), and 2.2 × 10−217(D).

Figure 2

Mean values and their 95% CIs of fasting levels of n-3 FAs (A and B), n-6 FAs (C and D), monounsaturated FAs (E and F), and saturated FAs and n-7 and -9 FAs (G and H) across the entire range of fasting and 2-h glucose categories in the cross-sectional METSIM Study cohort. P values (from ANOVA post hoc tests) indicating statistical significance with respect to the reference category (fasting or 2hPG <5.0 mmol/L) are coded as follows: *P < 0.05, **P < 0.01. P values for trends, adjusted for age and BMI, were as follows: 1.0 × 10−3 (A), 1.9 × 10−4 (B), 4.3 × 10−63 (C), 1.2 × 10−146 (D), 3.9 × 10−41 (E), 2.0 × 10−78 (F), 5.1 × 10−52 (G), and 6.1 × 10−132 (H).