Table 2.

Results of meta-regression analysis: Economic and socio-demographic factors associated with the obesity-undernutrition ratios based on reported prevalence in 29 studies^*

Predictors	Beta	Standard Error	p value
Sex
Male (ref, ND = 11)
Female (ND = 31)	−1.77	1.96	0.37
Male and Female (ND = 29)	1.24	2.15	0.56
Age group
≤ 8 years old (ref, ND = 27)
9–29 years old (ND = 18)	1.97	1.89	0.30
≥ 30 years old (ND = 26)	7.09	2.15	0.002
Country income level
High (ref, ND = 13)
Upper-middle (ND = 36)	−7.67	2.18	< 0.001
Lower-middle and low (ND = 22)	−10.78	2.59	< 0.001
Time trend
Data collection before 2000 (ref, ND = 18)
Data collection since 2000 (ND = 53)	5.24	1.53	< 0.001

ND: number of data points.

^*71 data points from 29 studies were analyzed using a Stata “Metareg” command with dummy variables of age group, sex, country income level, and secular trend.

Age, country income, and secular trend were categorized at the points of slope change into two/three groups. These categories were subsequently used in regression models given their non-linear relationship of age with the obesity-undernutrition ratio.