Macao survey of 14,311 measurements links adult leg length to childhood GDP per capita
Across four survey waves in 2005–2020, leg length of mainland-born adults rose by 1.02 cm per doubling of childhood GDP per capita; the study is observational.

Researchers analysed four waves (2005–2020) of Macao's citywide repeated cross-sectional physical fitness survey and reported the results in Frontiers in Public Health. The dataset includes 14,311 measurements from at least 13,874 individuals aged 20–59 years. Standing and sitting heights were measured, and leg length was derived from them. frontiersin.org
Frontiers in Public Healthfrontiersin.orgOpen source ↗
The authors start from the fact that relative leg length is fixed by the end of growth, so adult proportions reflect the childhood growth environment. Mainland-born adults grew up amid a 10-fold increase in GDP per capita in their region of origin. As a contextual index, the team used GDP per capita at age 5 and examined it among mainland-born adults, among those who arrived as adults, and in both origins pooled.
Leg length and childhood GDP
Among mainland-born adults, leg length increased by 1.02 cm per doubling of childhood GDP per capita (95% confidence interval [CI]: 0.81–1.23). Among those who arrived at age 18 or later, the figure was 0.66 cm (0.30–1.02). With both origins pooled and a flexible birth-year adjustment, it was 0.51 cm (0.12–0.89).
The association was nearly linear on the log scale. It was larger in men than in women: 1.43 versus 0.83 cm, with an interaction p < 0.001. frontiersin.org
Frontiers in Public Healthfrontiersin.orgOpen source ↗
Birthplace and education
Birth cohort, birthplace, education and occupation were examined in the full sample, adjusting for sex and age. The deficit among mainland-born adults was 1.17 cm at its widest, in the 1970s cohorts. It closed to a non-significant 0.05 cm among those born from 1990 onward, as the income gap narrowed.
More educated adults had longer legs, by 0.556 cm per step of a six-level scale (0.484–0.629). This educational difference showed no significant narrowing across cohorts: −0.023 cm per step per decade (95% CI: −0.085 to 0.040). The authors note that the interval cannot exclude a modest narrowing.
The authors stress that these are observational, population-level associations. They do not establish causation and do not indicate how improvements were distributed among families. frontiersin.org
Frontiers in Public Healthfrontiersin.orgOpen source ↗


