Research Article

Can secular trends in child growth be estimated from a single cross sectional survey?

BMJ 1992; 305 doi: (Published 03 October 1992) Cite this as: BMJ 1992;305:797
  1. C. A. Monteiro,
  2. A. M. Torres
  1. Nutrition Unit, World Health Organisation, Geneva, Switzerland.


    OBJECTIVE--To formulate and evaluate a new method to measure secular trends in child growth based on the cross sectional collection of heights of children and young adults. DESIGN--Trends in child growth obtained from comparison of two national surveys made with an interval of 15 years were compared with estimates obtained from comparison of height deficits of children and young adults in the more recent survey. SETTING--Brazil. SUBJECTS--Random sample of children (6 and 7 years old) and young adults (21 and 22 years old) living in Brazil in 1974 and 1989 (a total of 23,271 subjects in 1974 and 5479 in 1989). MAIN OUTCOME MEASURES--Increments in average heights of 6 and 7 year old children in a 15 year period. RESULTS--Mean height of 6 year old children increased 4.0 cm (boys) and 3.3 cm (girls) from 1974 to 1989. Similar results were obtained by subtracting, in the 1989 survey, mean height deficits found at ages 21 and 6 (3.8 cm for males and 3.5 cm for females). Positive changes in the mean height of 7 year old children could also be predicted by subtracting, in the 1989 survey, height deficits found at ages 7 and 22. CONCLUSIONS--Findings of this study support the hypothesis that secular trends in child growth can be estimated by comparing height deficits observed in children and young adults.