To read the full abstract: Diabetes Care. 2019;42(9):181623. doi: 10.2337/dc19-0430
Short summary: In this longstanding cohort study (n =1718), high intraindividual variability over time in BMI and in HDL-C during childhood, independent of their mean levels, conferred an increased risk of later-life diabetes.
Comment: Intraindividual variabilities in cardiovascular disease risk factors (CVRFs) such as BMI, blood pressure and the atherogenic lipids profile, have recently triggered interest as potential novel markers that can contribute to risk prediction, in addition to the absolute levels of CVRFs. Among adults, long-term follow-up data show consistent associations of higher variability in BMI, blood pressure and lipids with future adverse outcomes such as cardiovascular disease events, cognitive dysfunction and mortality. However, in adults, CVRF variability itself could result from common confounding risk factors such as older age, pre-existing comorbidities and medications. Childhood, a period during which few confounding risk factors or comorbidities exist, is the perfect time to assess the impact of intraindividual variabilities on CVRFs.
The Bogalusa Heart Study (BHS) is unique in having repeatedly measured and prospectively collected measurements of CVRFs from childhood through young adulthood to midlife. Between 1973 and 2016, 9 cross-sectional surveys of children aged 419 years and 11 surveys of adults aged 2058 years, who had been previously examined as children, were conducted approximately every 34 years.
During the 20.5-year follow-up period, 133 of 1718 participants developed diabetes. Increased variability in BMI and in HDL-C in childhood were associated with greater diabetes risk. This applied to only high BMI variability or only high HDL-C variability. Interestingly, other measures of the metabolic syndrome such as systolic/diastolic blood pressure, total cholesterol, triglycerides and LDL-C were not significantly associated with diabetes.
The authors suggest a number of pathophysiological mechanisms that could underlie the observed associations of intraindividual childhood BMI or HDL-C variability with diabetes. Among these possibilities, increased variabilities in BMI and HDL-C could reflect unhealthy lifestyle, or they could reflect an essential trait of responses or adaptations to alterations in neuroendocrine signals. Previous data demonstrated that fluctuations in BMI causes an unfavourable body composition characterized by more fat mass and less lean mass. In summary, dysfunction in CVRF homeostasis impacts the development of diabetes later in life.