Tracking of body mass index from 7 to 69 years of age

J Aarestrup, L G Bjerregaard, M Gamborg, L Ängquist, A Tjønneland, K Overvad, A Linneberg, M Osler, E L Mortensen, F Gyntelberg, R Lund, T I A Sørensen, J L Baker

65 Citations (Scopus)


BACKGROUND: Heavy children have an increased risk of being overweight young adults. Whether this risk remains in late adulthood is not well-understood. We investigated body mass index (BMI; kg m(-2)) tracking from childhood to late adulthood.

METHODS: From the Copenhagen School Health Records Register, 72 959 men and 25 252 women born between 1930 and 1989 with BMI values at 7 and/or 13 years and as adults were included. Using a meta-regression approach, age- and sex-specific partial correlation analyses and logistic regressions were performed.

RESULTS: Correlations between BMI at 7 years and young adult ages (18-19 years) were r=0.55 for men and r=0.55 for women. At late ages (60-69 years) these were r=0.28 for men and r=0.26 for women. The correlations did not differ by birth years. Compared with normal-weight 7-year-olds, overweight children had a higher odds of overweight at 18-19 years; odds ratio (OR)=14.02 (95% confidence interval (CI): 12.14-16.19) for men and 10.46 (95% CI: 4.82-22.70) for women. At ages 60-69 years ORs were 5.46 (95% CI: 0.95-31.36) for men and 1.61 (95% CI: 0.83-3.15) for women. Correlations and ORs were stronger at age 13 years than age 7 years as expected, but the overall patterns were similar.

CONCLUSIONS: BMI tracking was weaker at late adult ages than at young adult ages. Although BMI tracks across the life course, childhood BMI is relatively poor at identifying later adult overweight or obesity at ages when chronic diseases generally emerge.

Original languageEnglish
JournalInternational journal of obesity (2005)
Issue number9
Pages (from-to)1376-83
Number of pages8
Publication statusPublished - Sept 2016


  • Journal Article


Dive into the research topics of 'Tracking of body mass index from 7 to 69 years of age'. Together they form a unique fingerprint.

Cite this