Mounting evidence supports an association between nonoptimal ambient temperatures (i.e., heat or cold) and risk of low birthweight (LBW) (<2500 g), while the effect of temperature variability (TV) is largely unknown. We aimed to quantify the association between TV and risk of LBW in Africa.

Data on birthweight in 37 countries during 1990–2020 were collected from the Demographic and Health Surveys program. We calculated overall, intraday, and interday TV during the entire pregnancy and each trimester using hourly temperatures at ∼ 9 km resolution from ERA5-Land. We employed generalized linear mixed logistic regression, with random effects for country and survey cluster, to quantify the association between LBW and three separate TV metrics.

In total there were 33,863 (10.2%) LBW births out of 333,618 records. We found a J-shaped association between TV and LBW. Compared to the reference TV where the lowest risk was observed, extremely high (97.5th percentile) overall, intraday, and interday TV during the entire pregnancy increased the odds of LBW birth by 37.3% (26.7–48.8%), 24.1% (16.4–32.3%), and 15.1% (6.9–24.0%), respectively. In total, 7.3% of all LBW births in Africa were attributable to elevated overall TV. These associations were observed in dry climate zones, but not in tropical or temperate zones.

Our study suggests an adverse impact of TV on the risk of LBW in Africa, according to three different TV definitions, underlining the significance of climate-health risk assessment in those most vulnerable to climate change.