Relationship between the geographical distribution of adolescent body size and photoperiod observed in Japan and China: a spatial analysis

Geographic analyses of Japanese children have shown a paradoxical trend: effective daylength – the duration of sunlight above a given illumination threshold – is negatively associated with height and positively with weight adjusted for height. These patterns suggest photoperiodic influences, possibly resembling thyroid hormone effects. To investigate whether similar associations exist in Han Chinese children, we analyzed province-level data from 2019 on average height and weight, using annual mean global solar radiation at each provincial capital as a proxy for effective daylength. We applied the regression model: Height = a₀ + a₁ × Weight − a₂ × Solar Radiation. Under normal physiological conditions, height and weight are typically proportional; thus, support for this model would imply solar radiation is linked to reduced height and increased weight. To assess regional variation, we used geographically weighted regression (GWR), which estimates location-specific coefficients. The results showed spatial heterogeneity: the weight coefficient was greater in western provinces, while the solar radiation coefficient tended to be smaller at higher latitudes. A global regression for provinces north of 30°N revealed statistically significant associations for 9-year-old boys. These findings suggest that the height – daylength and weight – daylength relationships observed in Japan may also exist in northern China, though weaker and more variable.