BACKGROUND

The aim of this study was to quantify the relationship between insulin resistance and haematological parameters from adolescence to adulthood.

METHODS

Participants from the EPITeen cohort were evaluated at 13, 17, and 21 years, through standardized procedures. A fasting blood sample was obtained, and insulin resistance was evaluated by the homeostatic model assessment (HOMA-IR). The cross-sectional association between HOMA-IR and haematological parameters at 21 years was quantified in 1671 participants by multivariate linear regression coefficients (β) and 95% confidence intervals (95%CI). For the longitudinal analysis (n = 496), trajectories of insulin and glucose were estimated using model-based clustering, and haematological parameters were compared according to trajectories using ANOVA.

RESULTS

At 21 years, after adjustment for BMI, positive associations (β [95%CI]) were found between HOMA-IR and red blood count (0.05 [0.03;0.07] in females; 0.02 [0.00;0.04] in males); and haematocrit (0.29 [0.12;0.46] in females; 0.21 [0.04,0.38] in males). In females, HOMA-IR was inversely associated with packed cell volume (PCV) (-0.35 [-0.66;-0.05]) and iron levels (-3.98 [-6.94,-1.03]) but positively associated with white blood cells (0.31 [0.19;0.43]) and platelets (7.66 [3.93;11.39]). In males, a higher HOMA-IR was significantly associated with higher haemoglobin (0.09 [0.03;0.16]). Regarding the longitudinal analysis, similar trends were found, but statistical significance was not reached.

CONCLUSIONS

Both longitudinal and cross-sectional analyses support the hypothesis that insulin resistance is associated with increased red blood cells count and haematocrit in young adults, even within normal ranges of insulin and glucose.