• Gender-assortative waist circumference;
  • subcutaneous;
  • visceral fat


What is already known about this subject

  • Body mass index (BMI) is gender assortative (mother–daughter, father–son) in contemporary families.
  • The impact of obesity on the metabolic health of contemporary children is nevertheless a correlate of their own weight, not that of their parents.
  • Waist circumference (WC; abdominal fat) is a better predictor of metabolic risk than BMI.
  • Abdominal fat has two components – visceral, which increases metabolic risk, and subcutaneous, which is neutral or even protects.

What this study adds

  • WC, like BMI, is gender assortative. Daughters derive their girth from their mother and sons from their father.
  • Subcutaneous fat accounts for the gender-assortative link with parental WC, but not the surrogate we used for visceral fat (WC|BMI).
  • Insulin resistance, on the other hand, is a function of WC|BMI but not of subcutaneous fat.
  • Given appropriate standards, WC|BMI may help refine the predictive value of WC for metabolic risk in children.


Body mass index (BMI) is reportedly gender assortative (mother–daughter, father–son) in contemporary children. We investigated the corresponding transmission of waist circumference (WC) and its implications.


We measured parental WC at baseline and WC, height, weight and para-umbilical skin-fold (USF) annually in their offspring from 5 to 15 years (n = 223 trios). Parents were deemed normal metabolic risk (NR) or high risk (HR) according to World Health Organization (WHO) cut-points for WC (mothers 80 cm, fathers 94 cm). The residual from WC adjusted for BMI (WC|BMI) was used as a surrogate for excess intra-abdominal fat, and its association with insulin resistance (HOMA2-IR) was sought.


WC and USF were both gender assortative, while WC|BMI was not. WC was greater by 1.62 cm (P < 0.05, confidence interval [CI]: 0.09–3.16) and USF by 0.37 cm (P < 0.01, CI: 0.19–0.56) among the daughters (but not the sons) of HR compared with those of NR mothers, and by 1.32 cm (P < 0.05, CI: 0.09–2.55) and 0.18 cm (P < 0.05, CI: 0.04–0.32), respectively in the corresponding father–son (but not father–daughter) pairings. No such differences could be demonstrated for WC|BMI. A standard deviation score 1(SDS) change in WC|BMI, independent of BMI, was associated with a 7.14% change in IR in girls (P < 0.01, CI: 1.76–12.80) and 8.02% in boys (P < 0.001, CI: 2.93–13.36), but there was no relationship between IR and USF.


The relationship of offspring WC to metabolic health and to parental size is complex. Subcutaneous abdominal fat is gender assortative but harmless, while intra-abdominal fat (its surrogate in this analysis) is unrelated to parental waist circumference, but metabolically harmful.