Long run height and education implications of early life growth faltering: a synthetic panel analysis of 425 birth cohorts in 21 low- and middle-income countries

Table 2 Multivariable Regressions for the Association between Cohort Average Under-5 HAZ and Adult Height

	[1]	[2]	[3]
VARIABLES	Fully Adjusted Pooled Sample	Fully Adjusted WOMEN	Fully Adjusted MEN
Under-5 HAZ	2.005***	1.878***	2.921***
Under-5 HAZ	(1.086–2.924)	(0.870–2.887)	(1.423–4.419)
Controls
Child Cohort Characteristics
Pct. urban at childhood	−0.581	− 0.754	0.400
Pct. urban at childhood	(−1.930–0.767)	(− 2.245–0.736)	(−1.787–2.587)
Pct. electricity at childhood	0.688*	0.575	1.165**
Pct. electricity at childhood	(− 0.0299–1.406)	(− 0.495–1.645)	(0.254–2.075)
Avg. mat. Educ. at childhood	0.0220	0.0820	−0.496***
Avg. mat. Educ. at childhood	(−0.288–0.332)	(−0.256–0.420)	(−0.783 - -0.209)
Log income at childhood	−0.577***	−0.523**	− 0.563
Log income at childhood	(− 1.008 - -0.146)	(− 0.991 - -0.0545)	(−1.481–0.354)
Adult Respondent Characteristics
Female (1 = yes)	−12.06***
Female (1 = yes)	(−12.65 - -11.47)
Urban (1 = yes)	1.168***	1.185***	0.930***
Urban (1 = yes)	(0.950–1.385)	(0.962–1.409)	(0.265–1.594)
Observations	33,998	30,414	3584
R-squared	0.385	0.167	0.125

*** p < 0.01, ** p < 0.05, * p < 0.1
Notes: The outcome variable in all regressions is adult height in cm. All models are estimated using ordinary least squares, with 95% confidence intervals presented in parentheses. Standard errors are clustered at the country-region-survey year level. Column 1 presents results for the full sample, while columns 2 and 3 present results for the subsample of women and men, respectively

ISSN: 1471-2458