Table 4

Associations between MEPI and ARI by child age and region

SubgroupARI, N (%) yesARI aOR (95% CI)*P value**
MEPI per child’s age†‡
 0–2 years (N=172 665 (44.5%))15 873 (9.2)1.07 (1.05 to 1.09) <0.001
 2–5 years (N=215 386 (55.5%))16 040 (7.4)1.04 (1.03 to 1.06) <0.001
MEPI per region‡§,¶
 South Asia (N=191 949 (49.4%))16 817 (6.38)1.05 (1.03 to 1.07) <0.001
 Southeast Asia (N=3840 (1.0%))363 (8.17)1.04 (0.94 to 1.14)0.47
 Western Africa (N=74 315 (19.2%))4991 (6.12)1.02 (0.99 to 1.05)0.18
 Eastern Africa (N=81 056 (20.9%))10 995 (11.76)1.07 (1.05 to 1.09) <0.001
 Central Africa (N=30 459 (7.8%))3845 (11.63)1.10 (1.04 to 1.16) 0.001
 Southern Africa (N=6432 (1.7%))830 (11.97)1.02 (0.98 to 1.07)0.37

  • Table is based on an imputed dataset.

  • P values are based on binary logistic regression models.

  • *Model adjusted for: child’s age, child’s sex, first-born, vitamin A supplementation, breastfeeding status, BCG vaccination, treatment sought for fever/cough, low birth weight, maternal age, maternal education, shelter index, drinking water source, household smoking, crowding, urban-rural status, rainy season and country.

  • †The interaction test MEPI-ARI by child age in years (continuous) was significant at p<0.001.

  • ‡For the subanalysis per region and child’s age, a significance level of p<0.006 was maintained to correct for multiple testing.

  • §The interaction test MEPI-ARI by country was significant at p<0.001.

  • ¶Regions with corresponding countries and survey years: South Asia (Afghanistan 2015, India 2015), Southeast Asia (Myanmar 2015), Western Africa (Benin 2017, Ghana 2014, Guinea 2018, Liberia 2013, Mali 2018, Nigeria 2018, Senegal 2017), Eastern Africa (Burundi 2016, Ethiopia 2016, Kenya 2014, Malawi 2016, Rwanda 2014, Zimbabwe 2015, Uganda 2016, Tanzania 2015), Central Africa (Chad 2014, Congo Democratic Republic 2013), Southern Africa (Lesotho 2014, Namibia 2013).

  • **P-values in bold are significant.

  • aOR, adjusted OR; ARI, acute respiratory infection; MEPI, Multidimensional Energy Poverty Index; n, total number of cases.