Mothers’ autonomy and childhood stunting: evidence from semi-urban communities in Lao PDR

Childhood stunting (height-for-age z-scores below − 2) remains a global health burden, with 156 million stunted children in the world in 2015 [1]. Stunting, a form of chronic undernutrition, is not only harmful to physical growth and cognitive development in the early childhood period but also causes detrimental effects throughout the entire life-cycle, including less educational attainment and lower labor productivity [2‐4]. Therefore, reduction in childhood stunting is important to both health and economic development.

The Lao People’s Democratic Republic (Lao PDR) has the highest prevalence of childhood stunting (44.2%) in the Indochina region. Despite Lao PDR’s higher gross national income per capita in 2014 (1660 USD) relative to Cambodia (1020 USD) and Myanmar (1270 USD), its child health and nutrition were inferior to those of neighboring countries [5]. This fact implies the existence of deep-rooted causes specific to Lao PDR that hamper the improvement of children’s health and nutrition.

A proper understanding of the determinants of childhood stunting is extremely important for designing effective policies and programs for children’s health and nutrition. The existing literature has examined the socioeconomic, political, and environmental determinants of childhood stunting in developing countries [6]. Many of these studies have applied the UNICEF’s analytical framework, which categorizes the main causes of child nutritional status into three factors: immediate, underlying, and basic [7]. “Mother”, one of the underlying causes, is widely considered to be vital in determining child nutritional status [8]. Previous studies have confirmed that mothers’ demographic factors and improved socioeconomic status, such as age, physical status, literacy, education, employment, and income, contributed to reducing the likelihood of childhood stunting [9, 10].

Additionally, previous empirical studies have shown that mothers’ autonomy plays an important role in determining child nutritional status [8, 9, 11‐13]. Here, “autonomy” refers to “the ability… to obtain information and use it as the basis for making decisions about one’s private matters and those of one’s intimates” [14] and “the control women have over their own lives—the extent to which they have an equal voice with their husbands in matters affecting themselves and their families” [15].

Several difficulties exist in measuring women’s autonomy for empirical analysis. First, women’s autonomy is multidimensional [15‐18], including social, economic, political, and psychological aspects [18]. Therefore, researchers must determine which aspects of women’s autonomy are positively or negatively associated with children’s health and nutrition. Second, results depend on the context of the study site. Most previous studies on women’s autonomy and child nutrition have been conducted in South Asia and Africa, where women generally have lower social status than men due to, among other factors, social structure. The findings of these studies cannot be generalized to societies in which women can afford higher social status. In Lao PDR, because the majority of the population follows a matrilineal system, women are believed to have relatively high status [19]. Several previous studies have examined the determinants of child nutrition in Lao PDR [20‐22], but little research has investigated the role of mothers’ autonomy on childhood stunting. Therefore, the purpose of this study was to measure mothers’ autonomy and assess its association with childhood stunting in Lao PDR.

We examined how mothers’ autonomy was associated with the likelihood of childhood stunting by examining original survey data gathered from four semi-urban communities in Vientiane Capital of Lao PDR. Specifically, we categorized mothers’ autonomy into five dimensions: self-efficacy, self-esteem, decision-making, freedom of mobility, and control over money. We then examined how each dimension was associated with the likelihood of childhood stunting by multivariate logistic regression analysis.

First, we confirmed that mothers’ higher self-efficacy for accessing health services was significantly associated with lower likelihood of childhood stunting. This finding suggests that if a mother has greater confidence in accessing health services, her child is less likely to be stunted. Similar results have been found in other countries. For example, mothers’ lower autonomy over medical treatment for their children during periods of illness was associated with negative child growth in an urban slum of Bombay, India [29]. In rural Chad, mothers’ higher influence on child feeding was positively related with children’s height-for-age [30]. However, we did not find any significant effect of mothers’ self-efficacy for borrowing money on childhood stunting in our sample. One possible explanation is that money borrowing was uncommon among mothers in our study site as we confirmed that 38% of the mothers had money to use. Thus, mothers’ confidence over money borrowing was unrelated to children’s health.

Second, we confirmed that mothers’ higher self-esteem, as measured by intolerance toward domestic violence by a husband, was significantly associated with reduced likelihood of childhood stunting. This result is consistent with that of a previous study in PDR using nationally representative data, which confirmed a statistically significant association between mothers’ intolerance toward domestic violence and children’s higher height-for-age [22]. In contrast, mothers’ attitude towards domestic violence had no association with childhood stunting in India [31] and Tanzania [32]. The above results thus suggest a positive association between maternal self-esteem and child linear growth in the matrilineal society of Lao PDR. Among individual components of mother’s self-esteem, a statistically significant difference between mothers with stunted children and those with non-stunted children was confirmed for “(3) Refuse sex with husband” (Table 1). Since this component is directly related to marital sexual relationship, the result implies that mother’s higher autonomy in terms of sexuality behavior with a husband would lead to the improvement of child growth.

Third, the degree of mothers’ sole decision-making was not associated with the likelihood of childhood stunting. This result suggests that child linear growth is not affected by who makes decisions in the household. Empirical studies in other countries have yielded mixed results on this point. For example, a cross-country study that analyzed DHS datasets from 12 developing countries found that mothers’ greater decision-making power was positively related to children’s height-for-age in India and Mali. Conversely, women’s decision-making power was negatively associated with height-for-age in Haiti and Malawi [33]. As decision-making variables, this study examined whether the woman had the final say in health care, making large household purchases, household purchases for daily needs, and visits to family or relatives. Regarding decision-making on healthcare, mothers’ higher autonomy in choosing their own health care was associated with improved height-for-age among children in Tanzania [32]. Similarly, whether mothers had the final say on their own health care was associated with lower likelihood of childhood stunting in rural Nepal [34]. However, no significant association was found between any of the household decision-making variables (cooking, healthcare, purchasing, and going out) and childhood stunting in India [31]. Laotian women in the matrilineal system are generally afforded greater decision-making power than those in other societies [35]. Our study findings showed that mothers’ sole decision-making was not associated with childhood stunting in this society.

Absence of significant effect of mother’s sole decision-making on the childhood stunting implies shared decision-making between wife and husband may be more beneficial with regard to child growth. To scrutinize this point, we constructed a composite score of the answer “Wife and husband jointly = 1, or 0 otherwise” for four different matters with regard to decision-making. Then, we used this composite score as an explanatory variable in the multivariate logistic regression and checked its effect on childhood stunting. The result shows that although direction of the estimated coefficient for shared decision-making was toward a reduction in the childhood stunting, its statistical significance was not confirmed. We need further investigation on what forms of decision-making between wife and husband would be more favorable in terms of child growth.

Fourth, mothers’ freedom of mobility was not a statistically significant determinant of childhood stunting. Laotian women in matrilineal society have less limitation in going out compared to those in South Asia, which likely affected this result. A study from India showed that mothers who did not need permission to go to the market were less likely to have a stunted child [31]. In contrast, no significant relationship between mothers’ mobility and children’s height-for-age was found in Egypt [36].

Fifth, the children in our sample were less likely to be stunted if their mothers had independent income. This result suggests that mothers who can use money as they wish are more likely to spend money on their child’s welfare. Similar results were found in South Asia. In India, whether mothers were allowed to set aside money for personal use was associated with reduced likelihood of childhood stunting [31]. In Nepal, mothers’ control over income was associated with higher height-for-age of their children [37]. Our study suggests that mothers’ financial resources are important for children’s growth regardless of women’s social status in different societies.

Finally, regarding mothers’ basic characteristics, we found that mothers’ height was significantly and positively associated with the likelihood of childhood stunting. This result is consistent with results from other developing countries. For example, children were more likely to be stunted if their mothers were lower height or underweight (Body Mass Index <18.5 kg/m²) in Kenya [38], Tanzania [39], and Sri Lanka [40]. Because little previous work has examined the association between mothers’ physical state and childhood stunting in PDR, our study will provide important input for future research.

Recently, an increasing number of studies have been conducted to examine the association between women’s autonomy and children’s health in developing countries. However, very few studies have examined the effects of women’s autonomy on children’s health for societies in which women have relatively higher status than men, such as Laotian matrilineal society. Therefore, our analysis provides new findings in this study area.

Our research had several limitations. First, the direction of causality from mothers’ autonomy to children’s health outcomes was not fully confirmed due to the nature of the cross-sectional data. Further study will be required to investigate this causality. Second, we could not identify the concrete reasons behind the association of mothers’ autonomy and childhood stunting. Future research should shed light on the mechanism behind this relationship. Third, our survey was conducted in restricted areas of Vientiane Capital, and thus the study results cannot be generalized to the entire Lao population. Despite these limitations, this study is the first to analyze the association between mothers’ autonomy and childhood stunting in Lao PDR, and it will thus serve as an important benchmark for further work.

We clarified which dimensions of women’s autonomy were associated with childhood stunting in Lao PDR. A closer examination of mothers’ autonomy will aid proper understanding of the determinants of childhood stunting.