Correction to: Nature https://doi.org/10.1038/s41586-023-06418-5 Published online 13 September 2023

The code that we used to fit meta-analyses for the proportion of children stunted by age in Fig. 3a and Extended Data Figs. 8–10 incorrectly used the total number of children in the denominator for incidence, which incorrectly included children no longer at risk of stunting (i.e., children who had become stunted at the previous age). In the “Onset of stunting in early life” section, in the text now reading “The percentage that experienced incident stunting onset between birth and 3 months ranged from 7% to 57% in each cohort and was 18% overall”, the percentages previously read “6%”, “47%” and “16%.”

We also made analogous corrections in the Supplementary information to figures of age-stratified stunting incidence. Following this correction, age-stratified stunting incidence levels were higher after birth, but trends remained similar overall. We revised the links in the Code availability section for the replication code to include these corrections. Additionally, we mislabeled the y axes of Fig. 4b and Extended Data Fig. 12 and their corresponding captions as “Incidence proportion” instead of “Proportion”. We have also corrected a typographical error in cited figures in the Methods “Linear growth velocity” section, and now cite Fig. 6 and Extended Data Fig. 14 in place of Fig. 5 and Extended Data Fig. 10. Additionally, the Reporting summary did not include an email address to contact for data requests, which is now updated. These corrections do not change the inferences we drew from the study. For comparison, original and revised figures are available in the Supplementary information. The updates appear in the HTML and PDF versions of the article.

Supplementary information is available in the online version of this amendment.