Epidemiology

Patterns and determinants of the double burden of malnutrition at the household level in South and Southeast Asia

Abstract

Objectives

Many developing countries currently face a double burden of malnutrition (DBM) at the household level, defined by the World Health Organization, as when a mother may be overweight or anemic, and a child or grandparent is underweight, in the same household. For the present study, we defined it as the coexistence of overweight or obesity in the mother, and at least one child under the age of 5 undernourished, within the same household. Although underweight has long been considered a major issue in South and Southeast Asia, overweight and obesity have also been identified as a growing problem. The main aim of this study was to assess the DBM at the household level and its major determinants in South and Southeast Asia.

Methods

We used population-representative cross-sectional data from the Demographic and Health Survey, conducted between 2007 and 2017, for eight South and Southeast Asian countries: Bangladesh, India, Nepal, Pakistan, Myanmar, Timor, Maldives, and Cambodia. Multivariate logistic regression was performed to identify the sociodemographic factors associated with DBM.

Results

A total of 798,961 households were included in this study. The pooled prevalence of overweight or obesity for the mother and stunted child was 10.0% (95% CI: 8.0.0–12.0), for OBM and wasted child, it was 7.0% (95% confidence interval (CI): 6.0–8.0), and for overweight or obese mother (OBM) and underweight child, it was 7.0% (95% CI: 6.0–8.0). The prevalence of any of these DBM coexistences was 12.0% (95% CI: 10.0–13.0) in all households. Statistically significant positive associations (p < 0.05) were found for each of these coexistences, and a higher age of the mother, mothers with a lower education, the richest household quintile, and households with more than four members.

Conclusion

It is imperative that “double duty” action policies are developed that tackle the DBM, rather than targeting undernutrition or overnutrition separately. The findings from this study suggest that the promotion of education for women may aid in tackling the double burden on a household level.

Introduction

In recent decades, the major public health concern of low- and middle-income countries (LMICs) was reducing underweight, due to the prevalence of malnutrition in these countries [1]. However, rapid economic transition, demographic changes, and urbanization in LMICs have led to overweight and obesity becoming more prevalent. Currently, these countries are experiencing a coexistence of underweight and overweight at the population, individual, and household level, a phenomenon referred to as the double burden of malnutrition (DBM). At the household level, this DBM most commonly occurs through the presence of overweight or obese mothers (OBM) and undernourished children in the same home, referred to as mother overweight/child underweight (MOCU) [2, 3]. The occurrence of DBM at these levels is concerning, as it is well known that both underweight and overweight have multifaceted consequences for survival, incidence of chronic diseases, healthy development, and the economic productivity of individuals, societies, and health care systems [4, 5], leading to the inclusion of a target to eradicate all forms of malnutrition worldwide within the Sustainable Development Goals [6].

There is large variation between countries in the occurrence of MOCU within households. A study based on 131 national surveys from LMICs found that the prevalence of MOCU within homes ranged from 1.8% of households in Ethiopia to 15.9% in Egypt [7]. In South and Southeast Asia, rising economic growth, demographic, and epidemiological transitions in the past decades have been accompanied by steady declines in the prevalence of child undernutrition and an increasing trend of overweight or obesity among women. For example, in Bangladesh between 2004 and 2014, child undernutrition decreased from 51 to 36% [8, 9], while the prevalence of overweight among women increased from 11.4 to 25.2% over the same period of time [10], pointing to increases in DBM and MOCU specifically in this region. In addition, a recent Lancet Commission report pronounced that taken together, Asian and African states spend about 4.11% of their gross domestic product on dealing with the costs of undernutrition [11].

These increases in MOCU, a key form of household-level DBM, are a major concern for both individuals and countries. If we are to intervene at the population, household, and individual levels, we must understand its determinants, so that we can target policies and interventions and tackle inequalities [12]. Previous studies in LMICs and South and Southeast Asian countries in particular have mostly focused on the nutritional status of the mother or child solely. Although some studies have identified a number of determinants of the DBM at the household level, including maternal age, household wealth status, and maternal education, in individual countries within South and Southeast Asia [13,14,15], none of them have taken a region-wide perspective.

The objectives of this study were (1) to examine the prevalence and trend of MOCU at the household level within South and Southeast Asian countries and the region, and (2) to identify the factors associated with MOCU within individual countries and the region as a whole.

Methodology

Data sources and procedures

We used population-representative repeat cross-sectional data from the Demographic and Health Surveys (DHS), conducted between 2007 and 2016, for eight South and Southeast Asian countries: Bangladesh, India, Nepal, Pakistan, Myanmar, Timor, Maldives, and Cambodia. Of these eight countries, multiple time-point data were available for Bangladesh, Nepal, Pakistan, Timor, and Cambodia. These surveys were conducted at 5-year intervals across LMICs, following standardized data collection and sampling procedures in all countries.

DHS data collection includes a stratified two-stage random sampling approach, consisting of a selection of census enumeration areas (EA) based on a probability proportional to size approach, followed by a random selection of households from a complete listing within the selected EA. Health- and welfare-related data were collected through interviewing women of reproductive age (15–49 years) and their children (0–5 years) within households. Voluntary written informed consent was obtained from the participant before interviewing. Ethical approval was given by ICF International (Calverton, MD, USA) institutional review board and by individual review boards within every participating country. The DHS data set can be obtained from the DHS program website following a data-sharing policy. Detailed descriptions of DHS sampling procedures, validation of questionnaire, and data collection methods are published elsewhere [16].

Outcome variables

We used the 2006 World Health Organization (WHO) growth standard to calculate the anthropometric z-scores to measure nutritional status of children. We classified children as undernourished using three categories: stunted = height-for-age z-score < −2 standard deviation, underweight = weight-for-age z-score < −2 standard deviation, and wasted = weight-for-height z-score < −2 standard deviation [17, 18]. We used Asian-specific BMI cutoffs used to classify the mother as underweight (<18.5 kg/m2), overweight (23.0 to <27.5 kg/m2), or obese (≥27.5 kg/m2) [19, 20].

To identify MOCU, we identified households in which the mother was overweight or obese and the child was undernourished, based on the three categories above. This resulted in three separate combinations of household MOCU: (1) OBM and underweight child = OBM/UWC, (2) OBM and stunted child = OBM/STC, and (3) OBM and wasted child = OBM/WSC. We also combined these for a final outcome variable in which within the household, there was an OBM and at least one of these three measures of undernutrition, i.e., underweight, stunted, or wasted = OBM/ANY. According to our inclusion criteria, a total of 798,961 households were included in this study from 1,017,819 households in the original data set. In total, 218,858 households were excluded because they did not have information about mother and child nutritional status, with both required for our study. Details on missing data are provided in Supplementary file (Supplementary Table 1).

Covariates

Factors related to both parental and household characteristics were identified from previous literature and included in the analysis. These included individual-level factors, such as place of residence (urban or rural), wealth index, mother’s education (no education, primary, or secondary/higher), mother’s age (categorized by 10-year age groups < 20 years, 20–29, 30–39, and 40–49 years), mother’s working status (yes or no), household members (≤4 household members, >4 household members), and exposure to mass media (one reads the newspaper, listens to the radio, or watches TV) as covariates.

We also included key household characteristics as covariates, which captured the sociodemographics of the surveyed households. Sociodemographic factors included geographical location of the household (rural or urban) and wealth index. Wealth index was constructed on the basis of durable goods and assets, by applying principal component analysis techniques [21]; we then stratified the households into the following five socioeconomic groups: poorest (Q1), poorer (Q2), middle (Q3), richer (Q4), and richest (Q5). In our analysis, we categorized household wealth index into three categories (Q1/Q2 = poor, Q3 = middle, and Q4/Q5 = rich).

Statistical analysis

We applied frequency distribution to summarize the categorical, continuous variables, and descriptive statistics. The variables found statistically significant (p < 0.05) in bivariate analysis were considered for multivariable analysis. Multivariate logistic regression was performed to identify the sociodemographic factors associated with the DBM. We reported odds ratio (OR) with 95% confidence interval (CI) and considered p < 0.05 as the cut-off level of statistical significance. Pooled analyses were adjusted for the fixed effects of country and survey year.

We conducted a multivariate logistic regression (Table 2 and Supplementary Tables) adjusted for survey year and other factors, such as place of residence, wealth index, mother’s education, age, mother’s working status, household member, and exposed to media. The complex sampling design with weighted sample was adjusted for in all analyses, and variations in errors due to complex sample design were controlled in all the analyses by using “svy” command; all analyses were conducted in STATA (version 14).

Results

Sociodemographic characteristics

Table 1 describes the sociodemographic characteristics of the participants, with 798,961 households included in this study. The mean age (±SD) of mothers at data collection in the total sample was 29.9 (±10.2) years, with this ranging within countries from a mean of 28.6 years in Timor (2009) to 32.6 years in Pakistan (2012). Overall, 28.5% of mothers had not received any formal education, and 10.7% had a higher education, with the prevalence of no education highest in Nepal and Pakistan, and the proportion of mothers with a higher education highest in India and Pakistan. Nearly one-third of the participants lived in urban areas (30.6%), ranging from 14.6% in the Maldives to 48.1% in Pakistan. The highest proportion of women in the richest quintile (29.7%) was found in Cambodia; the lowest was found in Timor (9.0%). Twenty-one percent of the families had never been exposed to any type of information media.

Table 1 General characteristics of the study population.

Overall prevalence

Figure 1 represents the prevalence of the DBM among mother–child pairs in South and Southeast Asian countries. The pooled prevalence of OBM and stunted child was 10.0% (95% CI: 8.0–12.0), OBM/WSC was 7.0% (95% CI: 6.0–8.0), and OBM/UWC was 7.0% (95% CI: 6.0–8.0). The coexistence of OBM and undernourished child (underweight, stunted, or wasted) was observed in 12.0% (95% CI: 10.0–13.0) of households. Using the most recent time points, the highest prevalence of all measures of DBM was found in Pakistan in 2017 (OBM/STC = 24%, OBM/WSC = 14%, OBM/UWC = 15%, and OBM/ANY = 28% (Fig. 1)). More than 70% of the total sample came from one study in India in 2016, demonstrating the prevalence of OBM/STC, OBM/WSC, OBM/UWC, and OBM/ANY at 8.0%, 7.0%, 7.0%, and 10.0%, respectively.

Fig. 1: Prevalence of household-level double burden of malnutrition among mother–child pair in South and Southeast Asian countries.
figure1

OBM/STC overweight or obese mother and stunted child, OBM/WSC overweight or obese mother and wasted child, OBM/UWC overweight or obese mother and underweight child.

Increasing prevalence of DBM was seen in all countries for which multiple time-point data were available. For example, in Bangladesh, in 2007, the prevalence of OBM/ANY was 6.0%, with this increasing to 12.0% in 2014 (Fig. 1). Similarly, in Nepal, the prevalence of OBM/ANY rose from 5.0% in 2006 to 10.0% in 2016.

Factors associated with household-level double burden of malnutrition

Table 2 reports the factors associated with the DBM among mother–child pairs at the household level using the most recent time-point data. Higher odds for households with OBM and underweight, stunted, or wasted child (OBM/ANY) were statistically associated with the oldest category mother’s age (40–49 years) compared to <20 years as a reference [OR: 3.60 (95% CI: 3.06–4.24), p < 0.05], highest wealth index group compared to the lowest [OR: 1.44 (95% CI: 1.35–1.53), p < 0.05], and mothers who were not working during the study period [OR: 1.14 (95% CI: 1.08–1.19), p < 0.05].

Table 2 Factors associated with household-level double burden of malnutrition among mother–child pairs in South and Southeast Asia adjusted with survey year.

Similar results were found for OBM/STC that showed statistically significant positive associations (p < 0.05) with mother’s age 40–49 years [OR: 3.31 (95% CI: 2.78–3.94), p < 0.05], richest household [OR: 1.41 (95% CI: 1.32–1.51), p < 0.05], and mothers who were not working during the study period [OR: 1.13 (95% CI: 1.07–1.19), p < 0.05].

Mother’s age of 40–49 years was also associated with OBM/UWC [OR: 3.90 (95% CI: 3.14–4.85), p < 0.05] and OBM/UWC [OR: 3.87 (95% CI: 3.12–4.80), p < 0.05]. Households from the highest wealth group had significantly greater odds of OBM/WSC [OR: 1.36 (95% CI: 1.26–1.47), p < 0.05] and OBM/UWC [OR: 1.36 (95% CI: 1.26–1.47), p < 0.05].

Once controlling for covariates, increasing trends in the odds of household-level OBM/STC, OBM/ANY, OBM/UWC, and OBM/ANY over time were found for Bangladesh, Nepal, Timor, and Cambodia (Supplementary Tables 2, 3, 4, 6, and 7). In Bangladesh (Supplementary Table 2), increasing odds of OBM/ANY were found from 2007 to 2011 [OR: 1.5 (95% CI: 1.28–1.82)] and 2014 [OR: 2.2 (95% CI: 1.82–2.57)]. The only other country for which three time points of data were available, Nepal, showed higher odds in 2016 and 2011 than 2006 for all household-level measures of DBM. Although confidence intervals suggested that these differences were not significant for 2011 [OR: 1.22 (95% CI: 1.00–1.53)], the higher odds in 2016 were OBM/ANY OR: 1.5 (95% CI: 1.2–2.0) (Supplementary Table 4).

Discussion

This study shows the first analysis of the relationship between the double burden of maternal overnutrition and child undernutrition in South and Southeast Asian countries, utilizing a country-wide representative sample. There are two major findings: first, the pooled prevalence for the region of OBM/STC was 10.0% (95% CI: 8.0–12.0), OBM/UWC was 7.0% (95% CI: 6.0–8.0), OBM/UWC was 7.0% (95% CI: 6.0–8.0), and OBM/ANY was 12.0% (95% CI: 10.0–13.0). Second, increasing wealth, older maternal age (40–49 years), and primary or below primary education were associated with increased odds of the household DBM.

Similar to our study, a previous study reported that the household prevalence of BDM in Bangladesh was 11% [15]. However, this study only focused on rural populations, whereas data used here included urban areas as well. In our study, we observed the highest prevalence of OBM/STC in Pakistan (24%) in 2017 and the lowest prevalence in Bangladesh in 2007 (5%). In our study, the pooled prevalence of OBM/STC was 12%, which was higher than that reported for many countries of Africa, Asia, and Latin America, including Ethiopia (1.8%), Senegal (3%), Chad (3.5%), Uganda (3.6%), Tanzania (4.1%), Rwanda (4.4%), Kazakhstan (2.5%), Uzbekistan (4.1%), Jordan (3.6%), Cambodia (4%), and Columbia (4%) [7]. Country-specific trend analysis found that household-level DBM increased over time in all countries in which multiple time-point data were available, despite decreases in childhood undernutrition. Although this study did not attempt to explain these increases, increasing maternal overweight/obesity found within this study over the same time period, would appear to be an important factor. A recent study also reported increases in women overweight/obesity in South and Southeast Asia, over recent decades [22].

In addition, we found that OBM/STC, OBM/WSC, OBM/UWC, and OBM/ANY were comparatively higher in those mothers who had no education. A study in Bangladesh suggests that secondary or higher education of mothers may have contributed effectively in reducing the risk of malnutrition in under-fives in the country [22]. A further study also indicates that discordant mother–child pairs were significantly less likely to occur in households in which the mother had some type of formal education relative to those in which the mother has no formal education [23]. This suggests that improved maternal education could contribute effectively to reduce the household-level DBM.

In our study, OBM/STC, OBM/UWC, OBM/WSC, and OBM/ANY showed statistically significant positive association (p < 0.05) with older mother’s age (40–49 years), higher income households, lower maternal education, and having more than four household members. Other studies also reported that the household level of DBM was associated with household income [24] and mother’s age [25]. In contrast to our findings, one recent study in Bangladesh reported that a maternal age of 21–25 years at first birth was significantly associated with higher odds of OBM/UWC, although our study considered maternal age at the time of survey rather than at first birth. At the same time, in our study, we found that older mother’s age (40–49 years) was associated with a higher risk of household-level DBM. This may be due to the increased risk of maternal overweight/obesity, a prerequisite for household DBM, at older ages found both in the present and previous studies [26].

A recent policy analysis by the WHO reported that policies tend to address only undernutrition or overweight or obesity but do not consider both [27]. According to our analysis, the household level of DBM is increasing, and country-level policies must change to reflect this, supporting recommendations for “double duty actions” recently proposed by the WHO.

The present study illustrates the current scenario of different forms of household-level DBM and its relation with important sociodemographic correlates in South and Southeast Asian countries. The use of nationwide representative samples from both urban and rural areas is a strength of this study. However, this study has some limitations. Because the study design was cross-sectional, establishment of causality between the identified factors and DBM was not possible. Another limitation is that we only analyzed the sociodemographic determinants of DBM. Determinants such as dietary intake and physical activity were not evaluated. In addition, heterogeneity was found in both the years of available data between countries and in the sample characteristics.

Conclusion

The current study indicates an increasing DBM at the household level in South and Southeast Asian countries. It is high time that malnutrition-prevention programs to tackle this problem were introduced at the household level. Maternal education was associated with a decreased risk of household DBM, and prevention approaches and policies in South and Southeast Asian countries should emphasize the promotion of women’s education as one component. At the same time, surveillance of DBM in these countries is crucial in developing, targeting, and evaluating intervention and policy approaches. The development of comprehensive surveillance systems to monitor the progress of DBM at household, regional, and national levels should be encouraged.

References

  1. 1.

    Hoque ME, Hasan MT, Rahman M, Long KZ, Al Mamun A. Double burden of underweight and overweight among Bangladeshi adults differs between men and women: evidence from a nationally representative survey. JPHN. 2017;20:2183–91.

    Google Scholar 

  2. 2.

    Doak CM, Adair LS, Monteiro C, Popkin BM. Overweight and underweight coexist within households in Brazil, China and Russia. J Nutr. 2000;130:2965–71.

    CAS  Article  Google Scholar 

  3. 3.

    Hanandita W, Tampubolon G. The double burden of malnutrition in Indonesia: social determinants and geographical variations. SSM-Popul Health. 2015;1:16–25.

    Article  Google Scholar 

  4. 4.

    Black RE, Victora CG, Walker SP, Bhutta ZA, Christian P, De Onis M, et al. Maternal and child undernutrition and overweight in low-income and middle-income countries. The Lancet. 2013;382:427–51.

    Article  Google Scholar 

  5. 5.

    Bloom D, Cafiero E, Jané-Llopis E, Abrahams-Gessel S, Bloom L, Fathima S, et al. The global economic burden of noncommunicable diseases. Geneva: Program on the Global Demography of Aging; 2012.

  6. 6.

    Robert KW, Parris TM, Leiserowitz AA. What is sustainable development? Goals, Indicators, Values, and Practice. 2005;47:8–21.

  7. 7.

    Dieffenbach SD. Are stunted child/overweight mother pairs a distinct entity or a statistical artifact? Demographic and Health Surveys, 1991–2009. Stein: Emory University; 2012.

  8. 8.

    NIPORT M, Macro O. Bangladesh demographic and health survey 2004. Dhaka: National Institute of Population Research and Training, Mitra and Associates, and ORC Macro; 2005.

    Google Scholar 

  9. 9.

    NIPORT M. Bangladesh demographic and health survey BDHS 2014: key indicators. Dhaka and Calverton, Maryland: National Institute of Population Research and Training (NIPORT), Mitra and Associates, and ICF International; 2014.

    Google Scholar 

  10. 10.

    Biswas T, Uddin MJ, Al Mamun A, Pervin S, Garnett SP. Increasing prevalence of overweight and obesity in Bangladeshi women of reproductive age: findings from 2004 to 2014. PloS ONE. 2017;12:e0181080.

    Article  Google Scholar 

  11. 11.

    Swinburn BA, Kraak VI, Allender S, Atkins VJ, Baker PI, Bogard JR, et al. The global syndemic of obesity, undernutrition, and climate change: the Lancet Commission report. Lancet. 2019;393:791–846.

    Article  Google Scholar 

  12. 12.

    Perez-Escamilla R, Bermudez O, Buccini GS, Kumanyika S, Lutter CK, Monsivais P, et al. Nutrition disparities and the global burden of malnutrition. BMJ. 2018;361:k2252.

    Article  Google Scholar 

  13. 13.

    Das S, Shah MF, Islam MS, Biswas T, Mahfuz M, Ahmed T. Prevalence and sociodemographic determinants of household-level double burden of malnutrition in Bangladesh. JPHN. 2019;22:1425–32.

  14. 14.

    Jehn M, Brewis AJE, Biology H. Paradoxical malnutrition in mother–child pairs: untangling the phenomenon of over- and under-nutrition in underdeveloped economies. Econ Hum Biol. 2009;7:28–35.

  15. 15.

    Oddo VM, Rah JH, Semba RD, Sun K, Akhter N, Sari M, et al. Predictors of maternal and child double burden of malnutrition in rural Indonesia and Bangladesh. The Am J Clin Nutr. 2012;95:951–8.

  16. 16.

    Schott W, Aurino E, Penny ME, Behrman JR. The double burden of malnutrition among youth: trajectories and inequalities in four emerging economies. Econ Hum Biol. 2019;34:80–91.

    Article  Google Scholar 

  17. 17.

    Kimani-Murage EW, Muthuri SK, Oti SO, Mutua MK, van de Vijver S, Kyobutungi C. Evidence of a double burden of malnutrition in urban poor settings in Nairobi, Kenya. PloS ONE. 2015;10:e0129943.

    Article  Google Scholar 

  18. 18.

    Kosaka S, Umezaki M. A systematic review of the prevalence and predictors of the double burden of malnutrition within households. Br J Nutr. 2017;117:1118–27.

    CAS  Article  Google Scholar 

  19. 19.

    WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet. 2004;363:157.

    Article  Google Scholar 

  20. 20.

    Pi-Sunyer X. Clinical guidelines on the identification, evaluation and treatment of overweight and obesity in adults-the evidence report. Obes Res. 1998;6:51S–210S.

    Article  Google Scholar 

  21. 21.

    Biswas T, Islam MS, Linton N, Rawal LB. Socio-economic inequality of chronic non-communicable diseases in Bangladesh. PloS ONE. 2016;11:e0167140.

    Article  Google Scholar 

  22. 22.

    Hasan MT, Soares Magalhaes RJ, Williams GM, Mamun AAJM. The role of maternal education in the 15‐year trajectory of malnutrition in children under 5 years of age in Bangladesh. 2016;12:929–39.

  23. 23.

    Jehn M, Brewis A. Paradoxical malnutrition in mother–child pairs: untangling the phenomenon of over-and under-nutrition in underdeveloped economies. Econ Hum Biol. 2009;7:28–35.

    Article  Google Scholar 

  24. 24.

    Sekiyama M, Jiang HW, Gunawan B, Dewanti L, Honda R, Shimizu-Furusawa H, et al. Double burden of malnutrition in rural West Java: household-level analysis for father-child and mother-child pairs and the association with dietary intake. Nutrients. 2015;7:8376–91.

  25. 25.

    Wong C, Zalilah M, Chua E, Norhasmah S, Chin Y, Nur’Asyura AS. Double-burden of malnutrition among the indigenous peoples (Orang Asli) of Peninsular Malaysia. BMC Public Health. 2015;15:680.

  26. 26.

    Biswas T, Townsend N, Magalhaes RS, Islam MS, Hasan MM, Mamun A. Current progress and future directions in the double burden of malnutrition among women in South and Southeast Asian countries. Curr Dev Nutr. 2019;3:nzz026.

  27. 27.

    World Health Organization. Double-duty actions for nutrition: policy brief. Geneva: World Health Organization; 2017.

Download references

Acknowledgements

We thank the MEASURE DHS Data Archive, ICF International, for providing access to the South and Southeast Asian Demographic and Health Surveys data.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Tuhin Biswas.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Biswas, T., Townsend, N., Magalhaes, R.J.S. et al. Patterns and determinants of the double burden of malnutrition at the household level in South and Southeast Asia. Eur J Clin Nutr (2020). https://doi.org/10.1038/s41430-020-00726-z

Download citation

Search

Quick links