"Fatness", "Fitness" and Cardiometabolic Health In Young People

INTRODUCTION: “Fatness” and “fitness” are recognised as important influences on cardiometabolic health in young people. Body mass index (BMI) is normally used to represent adiposity, whilst cardiorespiratory fitness (CRF) is usually presented as V̇O2MAX scaled to total body mass. The aim of the current thesis was to assess whether using different measures to represent “fatness” and “fitness” influences the relationships between these factors and cardiovascular disease (CVD) risk in young people. 

STUDY ONE: A systematic review (n=994,595) and meta-analyses (n=18,526) compared the efficacy of using BMI, waist circumference (WC) and waist-to-height ratio (WHtR) for indicating CVD risk in 11-19 year olds. In meta-analyses BMI was a strong indicator of systolic blood pressure (mean difference with 95% confidence intervals, normal weight versus overweight/obese: -6.61mmHg, -8.71 to -4.90), diastolic blood pressure (-3.87mmHg, -4.90 to -2.84), triglycerides (-0.19mmol·L-1 , -0.24 to -0.14), high-density lipoprotein cholesterol (0.10mmol·L-1 , 0.06 to 0.14) and insulin (-30.17pmol·L-1 , -43.12 to -17.21); but not total cholesterol, low-density lipoprotein cholesterol or glucose. WC and WHtR were strong indicators of the same risk factors, though ≤2 studies were available for each comparison for these adiposity indices. Narrative synthesis results supported these outcomes. There was however insufficient evidence to judge which adiposity index was most efficacious for assessing CVD risk in young people. STUDY TWO: A prospective, cross-sectional study with 18-20 year olds (n=64). CRF, body composition (dual-energy X-ray absorptiometry) and CVD risk (blood pressure, blood lipids, glucose, insulin and inflammatory markers) were measured. Agreement was relatively poor between CRF scores when different scaling methods were used (k≤0.60 for most, “moderate” agreement). SBP and metabolic risk score were significantly associated with absolute CRF (r=0.379 and 0.461 respectively). Leptin was strongly associated with absolute CRF (r=-0.602) and CRF scaled to total body mass (r=-0.602). CRF was more strongly associated with most CVD risk factors when scaled to a method that included a measure of body fat. These relationships were removed when adiposity was controlled for. The method of scaling CRF had a large influence on the classification of CRF and the relationship of “fitness” with CVD risk in young people, so should be considered carefully. 

STUDY THREE: A retrospective analysis of mostly obese 8-18 year olds (n=426). CRF, body composition (air-displacement plethysmography) and CVD risk were assessed. There was relatively poor agreement between CRF scores when different scaling methods were used (k≤0.60 for most, “moderate” agreement) and between adiposity measures (BMIsds v WCsds mean difference=0.27sds, 95% limits of agreement ±0.69sds). Adiposity was a strong, independent predictor of CVD risk; but there was little difference in the predictive ability of BMI, WC and WHtR. CRF was more strongly associated with CVD risk when scaled to methods influenced by body fatness, with relationships attenuated when adiposity was controlled for. No CRF measure added to the predictive ability of adiposity measures, thus there was no evidence “fitness” should be considered alongside “fatness” when assessing young people’s CVD risk. 

CONCLUSIONS: There was no clear difference in the efficacy of different adiposity measures for indicating CVD risk, thus the findings do not suggest BMI should be replaced by WC or WHtR as a marker of “fatness” in young people. The observed relationships between “fitness” and CVD risk were due largely to an influence of body fatness on the CRF score as a result of inappropriate scaling. There was also no support for including any measure of CRF alongside adiposity indices for predicting CVD risk in young people. Previous research concluding an independent relationship between CRF and CVD risk without considering potential mathematical coupling with adiposity, due to the method of scaling, should therefore be reconsidered.