Research | Open Access | Volume 9 (2): Article 78 | Published: 19 May 2026
Menu, Tables and Figures
Table 1: Year of survey and number of surveyed children under five years of age by country
Table 2: Proportion (%) of malnourished children according to pregnancy planning by country
Table 3: Results of logistic regressions on the relationship between pregnancy planning and malnutrition by country (Odds ratio with 95% confidence interval)
Sibiri Clement Ouédraogo1,&, Moussa Bougma2
1National Institute of Statistics and Demography, Ouagadougou, Burkina Faso, 2Higher Institute of Population Sciences, Joseph Ki Zerbo University, Burkina Faso
&Corresponding author: Sibiri Clement Ouédraogo, National Institute of Statistics and Demography, Ouagadougou, Burkina Faso, Email clementouedraogo100@gmail.com
Received: 11 Mar 2025, Accepted: 15 May 2026, Published: 19 May 2026
Domain: Maternal and Child Health
Keywords: Planning, pregnancy, mistimed, unwanted, malnutrition, sub-Saharan Africa
©Sibiri Clement Ouédraogo et al. Journal of Interventional Epidemiology and Public Health (ISSN: 2664-2824). This is an Open Access article distributed under the terms of the Creative Commons Attribution International 4.0 License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Cite this article: Sibiri Clement Ouédraogo et al., Pregnancy planning and child malnutrition in Sub-Saharan Africa: A multi-country DHS analysis. Journal of Interventional Epidemiology and Public Health. 2026; 9(2):78. https://doi.org/10.37432/jieph-d-25-00065
Introduction: Few studies have assessed the impact of pregnancy planning on malnutrition in sub-Saharan Africa, and those carried out in other contexts come up with rather divergent results. This study therefore analysed the relationship between pregnancy planning and child malnutrition in sub-Saharan Africa while highlighting the disparities that may exist between countries and sub-regions.
Methods: The latest data from the Demographic and Health Surveys of 34 countries in sub-Saharan Africa were mobilised, and bivariate and multivariate descriptive methods through binomial logistic regressions were used for the analysis.
Results: The results indicate that pregnancy planning is a key factor in stunting and underweight in sub-Saharan Africa. Compared to children whose birth was planned, those whose birth was mistimed or unwanted have, respectively, 13.4% (aOR=1.134, 95%CI: 1.086 – 1.184) and 15.5% (aOR=1.155, 95%CI: 1.070 – 1.246) more at risk of stunting. They are also 12.2% (aOR=1.122, 95%CI: 1.069 – 1.177) and 23.3% (aOR=1.233, 95%CI: 1.133 – 1.342) more likely to be affected by underweight, respectively, compared to children whose birth was planned. On the other hand, wasting, which is a sign of an unsuitable diet in the very recent period, is not influenced by pregnancy planning.
Conclusion: These results indicate that the influence of pregnancy planning on malnutrition in sub-Saharan Africa is confined to its chronic forms (stunting and underweight) and underscores the need to integrate family planning into nutritional policies to reduce long-term child malnutrition.
Reducing malnutrition remains a public health priority in many countries in sub-Saharan Africa. Malnutrition can be defined as being stunted (small height for age), wasted (low weight for height) or underweight (low weight for age). Its prevalence remains very high in sub-Saharan Africa. Of the total number of 768 million undernourished people in 2021, more than one-third (261 million) live in sub-Saharan Africa. The proportion of undernourished people is 9.8% worldwide. This prevalence is higher in sub-Saharan Africa (23.2%) compared to other regions of the world, with intra-regional disparities: 9.2% in Southern Africa, 13.9% in West Africa, compared to 29.8% in East Africa and 32.8% in Central Africa [1].
About 200 million children under 5 suffer from stunting or wasting. Food poverty ‒ the inability to access and consume a nutritious, diverse diet ‒ affects 181 million children under 5 in its most severe form and remains the main driver of child malnutrition, increasing risks of mortality, poor growth and development [2].
In the face of this scourge, it is essential to explore the multiple underlying factors that influence child nutrition. One potential factor remains pregnancy planning. Indeed, unwanted or mistimed pregnancies have been associated with more adverse measures of health and well-being for parents, infants, and children [3]. A planned pregnancy allows parents to prepare financially, emotionally, and nutritionally, reducing the chances of food insufficiency and inadequate care. Conversely, an unwanted pregnancy can put children at increased risk of malnutrition due to insufficient access to food resources, inadequate birth spacing, or limited coverage of maternal and child nutritional needs.
In the Asian context, several studies indicate a significant influence of pregnancy planning on early childhood measures that provide information on the child’s nutrition status. Using longitudinal data, a study shows that in Andhra Pradesh, India, children who were unwanted at birth have height-for-age scores that are significantly lower than those of children who were desired at birth [4]. Nationwide, unwanted children are more likely than desired children to be stunted (OR = 1.3) [5]. Also in Bangladesh, children who were undesired at conception are at high risk of stunting (OR=1.4), wasting (OR=1.4) or underweight (OR=1.3) compared to children whose birth was desired [6]. This result is also observed in Nepal where newborns born from unwanted pregnancies were more likely to remain stunted (OR = 1.25; 95% CI = 1.00-1.56) [7].
In other contexts, however, the results remain mixed. In Bolivia, children aged 12 to 35 months from mistimed or unwanted pregnancies are about 30 per cent more likely to be stunted than children from unwanted pregnancies [8]. In Peru, child growth is significantly influenced by pregnancy planning, but no effect was found in the other countries (Bolivia, Egypt and Kenya)[9]. Also in low- and middle-income countries, no significant impact of pregnancy planning was found on stunting, underweight and wasting [10]. Moving in the same direction, a prospective study conducted in northern Malawi shows that the risk of stunting in children does not vary significantly depending on whether the birth is desired or not [11].
Overall, there are very few studies that have assessed the impact of pregnancy planning on malnutrition in Africa sub-Saharan studies and those conducted in other contexts lead to divergent results. While the influence of pregnancy planning on malnutrition is well established in certain contexts, particularly in Asia, its transposability in sub-Saharan Africa remains uncertain, in particular because of the cultural and socio-economic dynamics that could moderate these effects. Indeed, in sub-Saharan Africa, social values and representations are central to influencing health behaviours. In many sub-Saharan societies, the child is perceived as a “divine gift” or a blessing, a conception that is part of religious or ancestral belief systems. This sacralized vision of motherhood and birth can generate a certain motivation for adequate care even when the pregnancy is not planned or desired.
In view of the above, this study aims to fill a scientific gap: the lack of pan-African data on the impact of mistimed and unwanted pregnancies on malnutrition. We hypothesize, on the one hand, an impact of pregnancy planning on stunting and underweight and, on the other hand, a lack of impact on wasting, which is a sign of an inadequate diet during the very recent period.
Data sources
The target population for this study is all children under 5 years of age at the time of the survey period. The data used are those of the Demographic and Health Surveys (DHS) carried out in sub-Saharan African countries. Conducted as part of the DHS program, the DHS aims to estimate a wide range of socio-economic, demographic, and health indicators at the level of the general population and sub-populations of women aged 15 to 49, children under 5, and men aged 15 to 59. In addition, data are collected using a similar methodology that promotes comparability of data across countries.
The selection of countries included in this analysis was based on the following criteria:
The latest data from each country are retained for the analyses. The data used relates to 34 countries (Table 1).
Sampling strategy and weighting
Sampling in DHS is generally designed to ensure adequate representativeness of key indicators at the national level and at the administrative level 1 (corresponding to regions or provinces). A stratified and two-stage area survey is applied. The primary sampling unit, also known as the cluster, is the enumeration area (EA). Clusters are drawn using a systematic draw with probability proportional to their household size, usually from the list of clusters established in the last census of population. After the first-degree draw, a mapping and enumeration of households is often necessary in order to update the household list. After an updated list of households in each cluster, a sample of households in each cluster is systematically drawn from one country to another in a second-degree manner, with equal probability.
Aggregating data from multiple DHS survey results in a large sample size, offering high statistical power to detect even weak associations. The levels of analysis used are national, subregional, and regional. For the subdivision by subregion, we used the regional division of the African Union.
The analyses take into account the complexity of the DHS sample design, using sampling weights to correct for unequal probabilities of selection. Two types of weighting are applied: individual weighting at the country level and corrected weighting for multi-country analyses. For data weighting at regional and subregional levels, there are essentially two options. The first is to resize the weight of each country so that it is proportional to the population of the country at the time of the survey. The second option is to resize so that the total weight is the same for each country. The first option has the problem that, in general, a large country, such as Nigeria, will completely dominate the results, so we opt for the second, which is used in some studies [12].
Variables
The dependent variables of this study are anthropometric measures that account for child growth indicators and malnutrition more generally. Anthropometry is commonly used to measure the nutritional status of children. The height and weight distribution of children under 5 years of age is compared to that of the WHO standard reference population. A child is said to be malnourished when he or she has values greater than two standard deviations below the median of the WHO Child Growth Standards [13]. To achieve this, the height-for-age, weight-for-height and weight-for-age indices are calculated.
Stunting (small height for age) is a measure of stunted growth that may be due to an unfavourable growth environment to which children have been exposed. Stunting undermines the physical and cognitive development of children, increases their risk of dying from common infections and predisposes them to overweight and non-communicable diseases later in life [1].
Wasting (low weight for height) is a measure of acute malnutrition. It is a sign of an unsuitable diet in the period immediately preceding the survey. Child wasting is a life-threatening condition caused by insufficient nutrient intake, poor nutrient absorption, and/or frequent or prolonged illness. Affected children are dangerously thin with weakened immunity and a higher risk of mortality [1].
Underweight (low weight for age) is a composite indicator that captures both chronic and acute undernutrition: it may reflect stunting, wasting, or a combination of both.
The main independent variable is pregnancy planning (m4 in the DHS database), which is captured through the following questions addressed to women aged 15-49 for each of their births that occurred in the last five years: When you became pregnant, did you want to be pregnant at that time? Did you want to have a child later or did you not want (or not) any more children? This variable has inherent limitations in its retrospective nature. Reporting an unwanted pregnancy is subject to criticism because it is reported by mothers after the birth of children and can therefore be subject to recall bias. However, in the absence of a reliable alternative, this variable can be used as a basis for highlighting the consequences of unintended pregnancies that can be refined by possible subsequent studies using longitudinal data. The “pregnancy planning” variable includes 3 modalities: Planned (pregnancy wanted at the time), Mistimed (pregnancy wanted later), and Unwanted (pregnancy not wanted, neither at the time nor later).
The other independent variables are socioeconomic and demographic characteristics that are generally known to be associated with the dependent variables. These are essentially the characteristics of the child, his mother, and the household in which he lives. We selected the place of residence (urban-rural), the household’s wealth quintile (very low, low, medium, high, very high), the mother’s level of education (none, primary, secondary or higher), the mother’s occupation (employed, unemployed), her marital status (in a union or not in a union), her age group (15-19 years, 20-34 years, 35-45 years), their exposure to the media (highly, low, or not exposed), birth order (1, 2-3, 4-5, 6 or more), and the interval with previous birth (first birth, minus 24 months, 24-47 months, 48 months or more). They will be used to control the relationship between the main independent variable and each dependent variable.
Analytical approach
Two types of analyses are implemented in this study: a descriptive analysis and an explanatory analysis. The descriptive analysis consists of an analysis of the association between pregnancy planning and each of the dependent variables through cross-tabulations and proportion tests in view of their qualitative nature. This descriptive analysis is carried out by sub-region and by country. The aim is to assess the effect of each of these variables on the relationships between pregnancy planning and dependent variables.
At the explanatory level, the relationship is analyzed in the presence of the other independent variables taken simultaneously through binomial or binary logistic regressions (the dependent variables each having two modalities). The objective was to identify the net effects of pregnancy planning on dependent variables. The analysis model can be illustrated by the following mathematical equation:
\[
Z_i = \ln \left( \frac{P_i}{1 – P_i} \right)
= \beta_0 + \beta_1 X_1 + \beta_2 X_2 + \cdots + \beta_n X_n
\quad \text{et} \quad
P_i = F(Z_i) = \frac{1}{1 + e^{-Z_i}}
\]
where Pi is the probability that the child has the trait studied (a trait that varies according to the dependent variable), Xj is the jth independent variable, and βj is the coefficient of the jth independent variable. Each coefficient measures the impact of a change in the related independent variable on the probability of having the trait studied. The quotient is called odds \( \frac{P_i}{1 – P_i} \)
The logarithmic transformation of this quotient is the logit z. The “odd ratios” or odds ratios (OR) are given by (ez). If this value is less than 1, children in category k are (1-OR)*100 percent less likely than children in the reference group to have the trait studied. An “odds ratio” greater than 1 means that this chance is OR (or OR*100 percent) times greater than the children in the reference category.
For each of the three dependent variables, 39 separate regression models are run: one regression model per country (34), one model per sub-region (4), and one model for sub-Saharan Africa as a whole. All explanatory variables were included simultaneously in the model to assess their net effect. Multicolinearity between these variables was assessed using the variance inflation factor criterion (VIF<10), and the model’s ability to distinguish between classes was assessed using the Roc curves. The data obtained were deemed satisfactory. In addition, to carry out these analyses, the Stata 16 software was used.
Ethical considerations
The Demographic and Health Surveys used have all received ethical approval from the relevant national committees in each country, as well as from ICF International’s Institutional Review Board (IRB). Survey protocols comply with international public health research standards. All participants in the EDS surveys gave their informed consent before participation. The investigators explained the objectives of the survey, the confidentiality of data, and the right to refuse to participate or to interrupt the interview at any time. In addition, this secondary analysis uses completely anonymised data where all personal identifiers have been removed. The data are publicly accessible via the DHS program website, in accordance with research data dissemination policies.
Descriptive results
Stunting is a common phenomenon in sub-Saharan Africa: one in four children (25.6%) suffers from it. The impact of pregnancy planning on this proportion remains unnoticeable. In fact, 25.6% of children whose pregnancies were planned suffer from stunting, compared to 25.4% and 26.2% respectively for children whose pregnancies were mistimed. At the bivariate level, the differences are not statistically significant at the 5% threshold. In addition, it appears that the proportion of stunted children remains variable by region. It is significantly lower in West Africa (22.3%), while it stands at nearly 27% in the other regions. Pregnancy planning has no impact on stunting in all subregions except Central Africa, where stunting affects more children with planned pregnancies (28.3%) than those with mistimed births (24.3%). Depending on the country, stunting is quite infrequent in Gabon (10.6%), The Gambia (12.4%) and Senegal (13.6%). However, it remains widespread in Burundi (49.3%), Mozambique (37.1%) and the DRC (37.1%). In the vast majority of countries, pregnancy planning has no impact on stunting. However, in Mozambique and Nigeria, stunting is less common in children whose pregnancies were mistimed or unwanted compared to children whose pregnancies were planned. In Lesotho and Namibia, on the other hand, stunting is paradoxically more common among children whose pregnancies were mistimed than those whose pregnancies were planned.
Much less common than stunting, wasting affects 6.1% of children in sub-Saharan Africa. Paradoxically, it remains more common among children whose birth was planned (6.5%) compared to those whose birth was mistimed (5.2%) or unwanted (5.0%). Explanatory analyses are still necessary in order to compare this result. In addition, the prevalence of wasting remains variable across sub-regions of sub-Saharan Africa. While West Africa remains the sub-region least affected by stunting, it remains the most affected by wasting, which affects 7.5% of children. The lowest prevalence is observed in southern Africa (4.1%). With regard to the impact of pregnancy planning on the occurrence of wasting in each of the subregions, it appears that in southern and eastern Africa, no significant incidence is observed. In contrast, in West Africa, children whose births were mistimed or unwanted are less affected by wasting than those whose births were planned. In Central Africa, only children whose births were mistimed are less affected by this phenomenon. Depending on the country, the prevalence of wasting varies from a minimum of 1% in Rwanda to 15.9% in Niger. In almost all sub-Saharan countries, pregnancy planning has no impact on wasting. Only Mozambique and Niger stand out, with lower prevalence of wasting among children whose birth was unwanted compared to those whose birth was planned.
In sub-Saharan Africa, one in five children (19.5%) is underweight, and this prevalence remains surprisingly higher among children with planned births (20.0%). Among children whose birth was mistimed or unwanted, the prevalence of underweight is estimated at 18.0% and 19.0%, respectively. These bivariate results, contrary to expectations, will be evaluated by further explanatory analyses. In addition, underweight is not very common in southern Africa (16.2%) compared to other sub-regions with prevalence rates above 19%. In southern and eastern Africa, pregnancy planning has no impact on underweight. Its influence can be seen in Central and West Africa, where underweight is less prevalent among children whose births were mistimed. Depending on the country, underweight reaches extreme values in Niger (41.7%), Burundi (36.5%) and Chad (34.0%). However, it is not very common in Gabon (6.8%), South Africa (7.8%) and Kenya (9.6%). In the vast majority of countries, pregnancy planning does not have an impact on underweight. However, compared to planned births, mistimed births are more affected by underweight births in Zimbabwe, while they are less affected in Senegal. At the same time, unwanted births are more affected by underweight in Namibia and Rwanda, while they are less affected in Chad (Table 2).
Explanatory results
Pregnancy planning has a significant impact on child stunting in sub-Saharan Africa, after controlling for other explanatory variables. Compared to children whose birth was planned, those whose birth was mistimed or unwanted were 13.4% (OR = 1.134, 95%CI: 1.086 – 1.184) and 15.5% (OR = 1.155, 95%CI: 1.070 – 1.246), respectively, more likely to be stunted (Table 3). In Southern Africa, both mistimed and unwanted pregnancies show increased risks of stunting (OR = 1.181 and 1.184), with extreme peaks in Lesotho (OR = 3.361 and 2.189). In West Africa, on the other hand, only unwanted pregnancies present a marked risk (OR = 1.212), with high values in Burkina Faso (OR = 2.279) and Gambia (OR = 2.440). In Central and East Africa, unwanted pregnancies are at the same risk of stunting as planned pregnancies. However, mistimed pregnancies are at increased risk of stunting in these two regions (OR = 1.123 in Central Africa and OR = 1.158 in East Africa), including Chad (OR = 1.229), Uganda (OR = 1.287) and Tanzania (OR = 1.319). These findings suggest that interventions targeting family planning may mitigate stunting, especially in countries where ORs exceed 1.2.
Pregnancy planning is not a determining factor in wasting in sub-Saharan Africa. Indeed, whether their births are planned, mistimed or unwanted, children have the same risk of being wasted. This result remains valid in all subregions. However, some countries have localised risks. In Tanzania, mistimed and unwanted pregnancies greatly increase the risk of wasting (OR = 2.102 and 3.620), a phenomenon unique in the database. Chad has an increased risk for mistimed pregnancies (OR = 1.255), while in Comoros, unwanted pregnancies paradoxically have a lower risk of wasting (OR = 0.476). These disparities could reflect contextual factors (access to food aid, weaning practices) or methodological biases (limited sample size). Wasting, therefore, appears to be less directly related to pregnancy planning, except in specific contexts. This calls into question the paradoxical results observed at the bivariate level, indicating that wasting is more frequent in children whose birth was planned (6.5%) compared to those whose birth was mistimed (5.2%) or unwanted (5.0%). It would therefore be a fallacious association.
Underweight is more common among children whose births were mistimed or unwanted in sub-Saharan Africa. In fact, all other things being equal, these children are respectively 12.2% and 23.3% more likely to be affected by wasting compared to children whose birth was planned. These results indicate a significant influence of pregnancy planning on underweight in a sub-Saharan context. There are marked regional variations. In southern Africa, both types of pregnancies have a higher risk (OR = 1.225 and 1.267), particularly in Lesotho (OR = 3.290 and 3.032) and South Africa (OR = 3.092 for mistimed pregnancies). In East Africa, the risks are significant only for mistimed pregnancies (OR = 1.168), with peaks in Tanzania (OR = 1.591) and Kenya (OR = 1.440). In Central and West Africa, mistimed births have the same risk of underweight births as those with planned births, while unwanted births are associated with a high risk (OR = 1.291 in Central Africa and OR = 1.223 in West Africa). The differences are greater in Gabon (OR = 1.576), DRC (OR = 1.614), Burkina Faso (OR = 2.096) and Liberia (OR = 1.832). These findings highlight that underweight, often linked to acute dietary deficiencies, is exacerbated by mistimed or unwanted pregnancies, likely due to limited resources or lack of psychosocial preparedness (Table 3).
Assessing the possibility of a higher risk of malnutrition in mistimed or unwanted births was the guideline of this study. Malnutrition was captured through the anthropometric indicators of stunting, wasting and underweight.
The results indicate that planning is a key determinant of stunting and underweight in sub-Saharan Africa. Compared to children whose birth was planned, those whose birth was mistimed or unwanted have respectively 13.4% (OR=1.134) and 15.5% (OR=1.155) more risk of stunting. In addition, underweight is more common in children whose births were mistimed or unwanted. In fact, all other things being equal, these children are 12.2% (OR=1.122) and 23.3% (OR=1.233) more likely to be affected by underweight, respectively, compared to children whose birth was planned.
The results obtained are similar to those observed in the Asian context, where mistimed and unwanted births are at greater risk of stunting or underweight [4–7]. On the other hand, they contrast with a study that found no significant impact of pregnancy planning on stunting and underweight in low- and middle-income countries [10]. They also run counter to the results achieved in northern Malawi, showing that the risks of stunting in children do not vary significantly depending on whether the birth is desired or not [11]. These discrepancies could be explained by the pooling of heterogeneous low- and middle-income countries in [10], which may mask the regional specificities of sub-Saharan Africa in terms of nutritional, epidemiological, and reproductive contexts. Methodological differences may also contribute, particularly the operationalization of pregnancy intention and the inclusion of variables lying on the causal pathway between intention and child nutrition (such as birth weight or antenatal care) which generates an overadjustment bias that mechanically attenuates the estimated effect.
The results obtained could be attributed to the inadequate healthcare provided for mistimed or unwanted pregnancies. In sub-Saharan Africa, analyses of DHS data have shown that mistimed and unwanted pregnancies are more often associated with inadequate prenatal care, as evidenced by lower rates of early prenatal visits and compliance with the recommended four prenatal visits [14]. Furthermore, health practices occurring immediately after childbirth (initiation of breastfeeding and postnatal care) are strongly influenced by pregnancy planning, whereas those occurring later (exclusive breastfeeding for 6 months, completion of the immunization schedule) are less so [15]. Other studies have also shown that mistimed or unwanted births are less likely to receive adequate breastfeeding and nutrition [16,17].
Also, these unintended births occur at times when mothers do not have the material means to take care of them. Indeed, unwanted or mistimed pregnancies can occur at any time, even when the mother is very busy at work or school, for example. The exhaustion associated with reconciling professional /educational obligations with an unintended pregnancy may reduce the ability to meet the child’s specific nutritional and health needs, even in the absence of voluntary intention. However, when it comes to planned pregnancies, mothers make sure that they get to the times in their lives when they have time to deal with it. In addition, the psychological stress associated with unintended pregnancies could also alter care practices.
As far as wasting is concerned, it is not influenced by pregnancy planning. Whether their births are planned, mistimed or unwanted, children have the same risk of being wasted. This result is in line with those of low- and middle-income countries [10] and runs counter to those observed in the Asian context [6]. Two complementary explanations may account for this specificity in relation to stunting and underweight.
The first explanation relates to the temporal nature of the three anthropometric indicators. Unlike stunting and underweight, which reflect cumulative nutritional deficits accrued over months or years and which are sensitive to prenatal trajectories, birth weight, and early-life feeding practices, wasting captures an acute and recent nutritional deficit, typically occurring in the weeks preceding measurement. It is predominantly driven by proximal and conjunctural factors (diarrheal episodes, respiratory infections, inadequate weaning, or seasonal food insecurity) rather than by the long-term socio-familial environment in which the child evolves. Pregnancy planning, by contrast, is a distal determinant whose influence operates primarily through prenatal investment, perinatal care, and early-life pathways. Its effect is therefore expected to be cumulative and to materialize on chronic indicators, whereas it is diluted, or even masked, on a punctual indicator dominated by recent shocks that affect all children irrespective of the intentional status of their conception.
A second, complementary explanation is statistical. The prevalence of wasting in sub-Saharan Africa is substantially lower than that of stunting or underweight, which mechanically reduces the statistical power to detect modest associations with distal determinants.
The absence of a detectable effect should therefore be interpreted not as evidence that pregnancy planning is irrelevant to children’s nutritional well-being, but as a reflection of the etiological and statistical specificities of acute malnutrition. Importantly, this equal risk of wasting does not imply the absence of overall nutritional vulnerability among children born from mistimed or unwanted pregnancies. It rather indicates that acute food and health shocks affect children uniformly, regardless of their planning status, while the disadvantages associated with unintended pregnancies accumulate over time and become visible on chronic indicators of malnutrition.
Finally, there are two limitations in this study. First, DHS data have limitations inherent in any retrospective, cross-sectional, and single-pass survey, including selection and omission biases. As such, the main independent variable (pregnancy planning) has inherent limitations in its retrospective nature. Secondly, the data used were not collected on the same date. Collection dates vary from country to country over the period from 2011 to 2022. This can lead to biases in the aggregation of data.
This study highlights the significant impact of pregnancy planning on child malnutrition in sub-Saharan Africa, based on a descriptive and explanatory analysis of data from 34 countries. The results clearly show that pregnancy planning is a key determinant of stunting and underweight in children. Mistimed or unwanted births are at higher risk of chronic malnutrition.
In contrast, the analysis reveals that pregnancy planning does not affect wasting, which is more related to short-term or cyclical factors, such as food shocks or episodes of acute illness. This distinction is essential to orient public policies and health programs towards interventions adapted to the nature of nutritional problems.
These results call for urgent targeting of family planning services in high-risk regions (southern/western Africa) and systematic integration of nutritional counselling into reproductive health programs, in particular for mistimed and unwanted pregnancies. Such interventions would break the cycle between unwanted pregnancies and malnutrition, contributing simultaneously to MDGs 2 (zero hunger) and 3 (mother and child health).
In conclusion, the results of this study support policies that promote family planning to reduce child malnutrition in sub-Saharan Africa sustainably. These actions could not only improve children’s health but also contribute to broader sustainable development goals in the region.
What is already known about the topic
What this study adds
| Country | Survey year | Number of surveyed children under 5 years of age |
|---|---|---|
| Southern Africa | – | 66 447 |
| Angola | 2015-16 | 13 549 |
| Lesotho | 2014 | 2 915 |
| Malawi | 2015-16 | 16 330 |
| Mozambique | 2011 | 10 291 |
| Namibia | 2013 | 4 818 |
| South Africa | 2016 | 3 389 |
| Zambia | 2018 | 9 398 |
| Zimbabwe | 2015 | 5 757 |
| Central Africa | – | 70 490 |
| Burundi | 2016-17 | 12 384 |
| Cameroon | 2018 | 9 043 |
| Chad | 2014-15 | 16 901 |
| Congo | 2011-12 | 17 228 |
| DRC | 2013-14 | 6 077 |
| Gabon | 2019-21 | 8 857 |
| East Africa | – | 76 037 |
| Comoros | 2012 | 3 023 |
| Ethiopia | 2016 | 9 939 |
| Kenya | 2022 | 18 698 |
| Madagascar | 2021 | 11 686 |
| Rwanda | 2019-20 | 7 737 |
| Tanzania | 2022 | 10 321 |
| Uganda | 2016 | 14 633 |
| West Africa | – | 136 288 |
| Benin | 2017-18 | 12 564 |
| Burkina Faso | 2021 | 11 757 |
| Gambia | 2019-20 | 7 883 |
| Ghana | 2022 | 9 005 |
| Guinea | 2018 | 7 233 |
| Ivory Coast | 2021 | 9 878 |
| Liberia | 2019-20 | 5 209 |
| Mali | 2018 | 9 221 |
| Niger | 2012 | 11 602 |
| Nigeria | 2018 | 30 547 |
| Senegal | 2019 | 5 845 |
| Sierra Leone | 2019 | 9 009 |
| Togo | 2013-14 | 6 535 |
| Table 2: Proportion (%) of malnourished children according to pregnancy planning by country | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Country/Sub-region | Stunting | Wasting | Underweight | |||||||||
| Total | Planned | Mistimed | Unwanted | Total | Planned | Mistimed | Unwanted | Total | Planned | Mistimed | Unwanted | |
| Southern Africa | 27.5 | 27.6 | 27.8ns | 26.1ns | 4.1 | 4.3 | 4.0ns | 3.5ns | 16.2 | 16.0 | 17.0ns | 15.4ns |
| Angola | 31.2 | 32.3 | 28.7ns | 29.8ns | 4.6 | 4.4 | 5.3ns | 3.5ns | 23.1 | 23.0 | 22.2ns | 28.3ns |
| Lesotho | 27.1 | 21.4 | 36.1*** | 26.9ns | 2.8 | 3.8 | 2.1ns | 1.6ns | 14.2 | 11.7 | 18.2ns | 13.9ns |
| Malawi | 29.7 | 29.7 | 29.8ns | 29.4ns | 2.5 | 2.1 | 3.2ns | 2.6ns | 15.7 | 15.1 | 15.8ns | 18.9ns |
| Mozambique | 37.1 | 38.3 | 29.9*** | 25.0*** | 5.2 | 5.3 | 5.0ns | 1.5*** | 19.9 | 20.5 | 18.2ns | 9.6*** |
| Namibia | 17.8 | 14.4 | 21.4** | 21.7ns | 8.1 | 8.0 | 6.9ns | 13.2ns | 18.1 | 16.1 | 19.0ns | 24.4* |
| South Africa | 20.4 | 15.4 | 24.7* | 23.3ns | 1.7 | 2.4 | 1.1ns | 1.3ns | 7.8 | 5.5 | 11.8ns | 6.4ns |
| Zambia | 28.1 | 27.0 | 29.8ns | 30.9ns | 3.8 | 3.6 | 4.3ns | 2.9ns | 16.5 | 15.9 | 17.3ns | 18.8ns |
| Zimbabwe | 21.2 | 19.9 | 23.4ns | 24.8ns | 3.6 | 3.5 | 3.4ns | 5.2ns | 11.8 | 10.4 | 13.9** | 16.6** |
| Central Africa | 27.2 | 28.3 | 24.3*** | 24.9ns | 5.7 | 6.2 | 4.6** | 5.1ns | 20.8 | 22.0 | 17.1*** | 19.9ns |
| Burundi | 49.3 | 49.4 | 47.7ns | 52.8ns | 4.3 | 4.1 | 4.6ns | 5.1ns | 36.5 | 36.1 | 36.0ns | 41.6* |
| Cameroon | 24.2 | 24.6 | 23.0ns | 22.4ns | 4.2 | 4.2 | 4.2ns | 5.4ns | 13.9 | 14.1 | 12.8ns | 13.6ns |
| Chad | 35.8 | 36.1 | 33.7ns | 27.9ns | 12.0 | 12.0 | 12.6ns | 10.6ns | 34.0 | 34.3 | 32.1ns | 23.0** |
| Congo | 19.3 | 18.2 | 22.7ns | 17.7ns | 5.0 | 5.2 | 4.3ns | 6.4ns | 15.5 | 15.0 | 16.8ns | 16.8ns |
| DRC | 37.1 | 39.0 | 32.5** | 31.1ns | 7.2 | 7.2 | 7.6ns | 6.1ns | 27.9 | 29.1 | 24.1* | 27.3ns |
| Gabon | 10.6 | 9.0 | 12.4* | 13.1ns | 3.0 | 3.1 | 2.5ns | 4.2ns | 6.8 | 6.3 | 6.6ns | 10.2ns |
| East Africa | 27.8 | 27.4 | 28.4ns | 29.6ns | 6.4 | 6.6 | 5.9ns | 6.0ns | 19.7 | 20.0 | 18.3ns | 21.3ns |
| Comoros | 26.1 | 25.1 | 28.5ns | 28.4ns | 10.2 | 10.3 | 11.0ns | 5.8ns | 20.4 | 19.4 | 22.8ns | 23.3ns |
| Ethiopia | 33.2 | 32.6 | 35.4ns | 34.2ns | 9.0 | 9.1 | 7.7ns | 11.0ns | 29.6 | 29.3 | 30.0ns | 31.3ns |
| Kenya | 19.3 | 18.3 | 20.1ns | 23.9** | 4.4 | 4.5 | 3.8ns | 5.8ns | 9.6 | 9.4 | 9.7ns | 10.6ns |
| Madagascar | 32.2 | 31.8 | 33.2ns | 38.6ns | 7.0 | 6.8 | 8.9ns | 8.5ns | 30.4 | 30.1 | 30.7ns | 38.2* |
| Rwanda | 27.0 | 25.9 | 28.9ns | 28.6ns | 1.0 | 1.0 | 1.3ns | 0.5ns | 10.8 | 9.9 | 10.8ns | 14.9** |
| Tanzania | 31.3 | 30.3 | 33.7ns | 35.1ns | 3.4 | 2.9 | 4.5ns | 6.7ns | 11.5 | 10.6 | 13.7ns | 14.2ns |
| Uganda | 22.7 | 21.5 | 24.8ns | 23.1ns | 3.3 | 3.1 | 3.5ns | 3.6ns | 13.3 | 11.8 | 15.5* | 15.9ns |
| West Africa | 22.3 | 22.5 | 21.0ns | 23.5ns | 7.5 | 7.8 | 6.4*** | 6.3* | 21.0 | 21.3 | 19.4** | 21.2ns |
| Benin | 25.9 | 25.9 | 25.6ns | 26.5ns | 4.4 | 4.3 | 5.3ns | 3.5ns | 21.4 | 21.0 | 22.7ns | 22.9ns |
| Burkina Faso | 22.0 | 21.8 | 21.5ns | 34.2ns | 13.4 | 13.3 | 13.1ns | 19.2ns | 19.1 | 18.7 | 21.7ns | 26.0ns |
| Gambia | 12.4 | 11.7 | 13.0ns | 29.6* | 5.0 | 5.3 | 3.2* | 7.3ns | 16.8 | 16.1 | 18.6ns | 25.7ns |
| Ghana | 18.4 | 18.8 | 17.6ns | 18.8ns | 8.0 | 8.2 | 7.3ns | 8.6ns | 14.1 | 14.3 | 14.3ns | 11.9ns |
| Guinea | 25.3 | 24.8 | 27.1ns | 34.2ns | 8.3 | 8.1 | 9.6ns | 7.4ns | 19.8 | 19.7 | 19.3ns | 26.1ns |
| Ivory Coast | 22.8 | 23.6 | 20.2ns | 20.5ns | 11.2 | 11.4 | 10.1ns | 11.5ns | 15.7 | 16.3 | 13.8ns | 15.1ns |
| Liberia | 22.5 | 23.3 | 21.1ns | 22.2ns | 2.9 | 3.0 | 2.4ns | 3.6ns | 15.1 | 14.2 | 14.6ns | 23.4ns |
| Mali | 22.3 | 22.2 | 21.9ns | 24.4ns | 8.8 | 8.7 | 9.3ns | 10.1ns | 23.5 | 23.2 | 25.2ns | 26.8ns |
| Niger | 36.2 | 36.6 | 32.2ns | 25.0ns | 15.9 | 16.2 | 12.8ns | 2.5*** | 41.7 | 42.0 | 39.0ns | 30.1ns |
| Nigeria | 31.8 | 33.0 | 22.6*** | 22.5*** | 6.1 | 6.3 | 4.8ns | 4.5ns | 26.6 | 27.0 | 22.9* | 23.0ns |
| Senegal | 13.6 | 13.9 | 11.1ns | 15.5ns | 8.2 | 8.5 | 6.5ns | 6.8ns | 19.2 | 19.9 | 14.1** | 17.9ns |
| Sierra Leone | 23.6 | 23.8 | 23.5ns | 18.9ns | 4.8 | 4.8 | 5.2ns | 2.8ns | 18.2 | 17.7 | 22.2ns | 15.8ns |
| Togo | 22.1 | 21.0 | 25.4ns | 23.0ns | 6.4 | 6.4 | 6.1ns | 7.0ns | 20.5 | 20.1 | 21.5ns | 21.8ns |
| Sub-Saharan Africa | 25.6 | 25.6 | 25.4ns | 26.2ns | 6.1 | 6.5 | 5.2*** | 5.0*** | 19.5 | 20.0 | 18.0*** | 19.0ns |
| Source: DHS | khi2 test significance: *** 1%; ** 5%; * 10%; ns not significant | ||||||||||||
| Table 3: Results of logistic regressions on the relationship between pregnancy planning and malnutrition by country (Odds ratio with 95% confidence interval) | ||||||
|---|---|---|---|---|---|---|
| Country/Sub-region | Stunting | Wasting | Underweight | |||
| Mistimed | Unwanted | Mistimed | Unwanted | Mistimed | Unwanted | |
| Southern Africa | 1.181*** [1.092 – 1.278] | 1.184** [1.024 – 1.368] | 1.012ns [0.855 – 1.198] | 1.069ns [0.799 – 1.432] | 1.225*** [1.115 – 1.346] | 1.267*** [1.074 – 1.494] |
| Angola | 1.690** [1.099 – 2.599] | 1.104ns [0.692 – 1.763] | 0.555ns [0.136 – 2.255] | 0.492ns [0.103 – 2.364] | 3.092*** [1.634 – 5.851] | 1.116ns [0.513 – 2.424] |
| Lesotho | 1.044ns [0.906 – 1.202] | 1.034ns [0.776 – 1.378] | 1.178ns [0.883 – 1.571] | 0.890ns [0.455 – 1.740] | 1.124ns [0.964 – 1.310] | 1.344** [1.007 – 1.795] |
| Malawi | 3.361*** [1.859 – 6.075] | 2.189** [1.097 – 4.368] | 0.919ns [0.0621 – 3.619] | 0.265ns [0.018 – 3.955] | 3.290*** [1.587 – 6.821] | 3.032** [1.253 – 7.335] |
| Mozambique | 1.021ns [0.884 – 1.178] | 0.979ns [0.778 – 1.234] | 1.372ns [0.924 – 2.036] | 1.282ns [0.662 – 2.481] | 1.079ns [0.903 – 1.289] | 1.321** [1.008 – 1.732] |
| Namibia | 1.035ns [0.883 – 1.214] | 1.040ns [0.781 – 1.385] | 1.196ns [0.866 – 1.651] | 0.453ns [0.173 – 1.182] | 1.244** [1.031 – 1.501] | 0.734ns [0.488 – 1.105] |
| South Africa | 1.508*** [1.105 – 2.058] | 1.499* [0.949 – 2.368] | 0.932ns [0.595 – 1.460] | 1.265ns [0.720 – 2.222] | 1.175ns [0.856 – 1.613] | 1.521* [0.975 – 2.375] |
| Zambia | 1.165*** [1.042 – 1.302] | 1.244* [0.987 – 1.567] | 1.194ns [0.924 – 1.543] | 0.694ns [0.376 – 1.280] | 1.139* [0.997 – 1.301] | 1.302* [0.993 – 1.708] |
| Zimbabwe | 1.069ns [0.906 – 1.262] | 1.229ns [0.930 – 1.624] | 0.882ns [0.605 – 1.287] | 1.548ns [0.895 – 2.677] | 1.177ns [0.957 – 1.446] | 1.631*** [1.175 – 2.263] |
| Central Africa | 1.123** [1.022 – 1.234] | 1.129ns [0.934 – 1.365] | 0.975ns [0.814 – 1.167] | 1.204ns [0.798 – 1.818] | 1.055ns [0.955 – 1.165] | 1.291** [1.053 – 1.582] |
| Burundi | 1.033ns [0.897 – 1.189] | 1.070ns [0.863 – 1.327] | 1.087ns [0.792 – 1.492] | 1.068ns [0.673 – 1.692] | 1.019ns [0.884 – 1.174] | 1.047ns [0.847 – 1.295] |
| Cameroon | 0.977ns [0.799 – 1.195] | 1.021ns [0.671 – 1.554] | 1.232ns [0.812 – 1.870] | 1.453ns [0.668 – 3.162] | 1.090ns [0.843 – 1.410] | 1.054ns [0.626 – 1.775] |
| Chad | 1.219* [0.997 – 1.491] | 1.053ns [0.628 – 1.768] | 0.845ns [0.572 – 1.249] | 1.290ns [0.563 – 2.953] | 1.088ns [0.872 – 1.356] | 1.250ns [0.739 – 2.111] |
| Congo | 1.228* [0.992 – 1.520] | 1.247ns [0.916 – 1.697] | 0.757ns [0.507 – 1.130] | 1.530ns [0.914 – 2.561] | 0.891ns [0.683 – 1.161] | 1.576*** [1.116 – 2.225] |
| DRC | 1.130* [0.992 – 1.287] | 1.208ns [0.941 – 1.550] | 1.014ns [0.812 – 1.265] | 0.797ns [0.508 – 1.249] | 1.061ns [0.924 – 1.217] | 1.614*** [1.252 – 2.080] |
| Gabon | 1.229** [1.047 – 1.444] | 0.746ns [0.484 – 1.151] | 1.255** [1.011 – 1.558] | 1.065ns [0.590 – 1.922] | 1.301*** [1.109 – 1.527] | 0.690ns [0.438 – 1.085] |
| East Africa | 1.158*** [1.046 – 1.281] | 1.092ns [0.947 – 1.259] | 1.013ns [0.833 – 1.233] | 0.958ns [0.728 – 1.260] | 1.168** [1.038 – 1.315] | 1.179* [0.991 – 1.402] |
| Comoros | 1.045ns [0.826 – 1.321] | 1.114ns [0.737 – 1.684] | 0.953ns [0.680 – 1.334] | 0.476** [0.230 – 0.985] | 1.095ns [0.848 – 1.413] | 1.081ns [0.693 – 1.687] |
| Ethiopia | 1.194*** [1.058 – 1.348] | 1.068ns [0.902 – 1.265] | 0.898ns [0.730 – 1.107] | 1.227ns [0.952 – 1.581] | 1.126* [0.994 – 1.275] | 1.166* [0.982 – 1.384] |
| Kenya | 1.169ns [0.966 – 1.416] | 0.942ns [0.702 – 1.264] | 1.269ns [0.873 – 1.846] | 1.577* [0.933 – 2.668] | 1.440*** [1.122 – 1.848] | 1.014ns [0.684 – 1.502] |
| Madagascar | 1.103ns [0.867 – 1.404] | 1.178ns [0.883 – 1.570] | 1.419* [0.956 – 2.106] | 1.090ns [0.670 – 1.773] | 1.052ns [0.824 – 1.343] | 1.183ns [0.889 – 1.575] |
| Rwanda | 1.287*** [1.082 – 1.531] | 1.204ns [0.906 – 1.601] | 0.993ns [0.655 – 1.504] | 1.284ns [0.667 – 2.472] | 1.362*** [1.103 – 1.682] | 1.403** [1.007 – 1.954] |
| Tanzania | 1.169ns [0.965 – 1.416] | 0.869ns [0.671 – 1.123] | 1.184ns [0.565 – 2.484] | 0.396ns [0.094 – 1.678] | 1.150ns [0.880 – 1.502] | 1.093ns [0.787 – 1.518] |
| Uganda | 1.319*** [1.074 – 1.620] | 1.369ns [0.878 – 2.136] | 2.102*** [1.301 – 3.397] | 3.620*** [1.466 – 8.939] | 1.591*** [1.202 – 2.107] | 1.589ns [0.869 – 2.906] |
| West Africa | 1.074* [0.995 – 1.160] | 1.212*** [1.052 – 1.397] | 0.904ns [0.800 – 1.021] | 0.946ns [0.755 – 1.185] | 1.058ns [0.978 – 1.144] | 1.223*** [1.055 – 1.418] |
| Benin | 1.212*** [1.072 – 1.371] | 1.210* [0.998 – 1.466] | 1.185ns [0.934 – 1.502] | 0.721ns [0.466 – 1.113] | 1.168** [1.029 – 1.326] | 1.256** [1.029 – 1.532] |
| Burkina Faso | 0.934ns [0.681 – 1.282] | 2.279** [1.209 – 4.298] | 1.043ns [0.717 – 1.516] | 1.555ns [0.756 – 3.201] | 1.222ns [0.891 – 1.677] | 2.096** [1.084 – 4.054] |
| Gambia | 0.834ns [0.640 – 1.087] | 0.861ns [0.506 – 1.466] | 0.792ns [0.560 – 1.118] | 1.216ns [0.623 – 2.374] | 0.805ns [0.595 – 1.090] | 0.970ns [0.541 – 1.740] |
| Ghana | 1.079ns [0.808 – 1.442] | 2.440*** [1.413 – 4.212] | 0.579** [0.344 – 0.977] | 1.219ns [0.477 – 3.116] | 1.259* [0.977 – 1.621] | 1.669* [0.958 – 2.908] |
| Guinea | 1.009ns [0.775 – 1.312] | 1.266ns [0.851 – 1.883] | 0.703* [0.481 – 1.025] | 0.772ns [0.439 – 1.356] | 1.056ns [0.791 – 1.411] | 0.939ns [0.589 – 1.499] |
| Ivory Coast | 1.333** [1.024 – 1.733] | 1.640** [1.008 – 2.669] | 1.151ns [0.780 – 1.698] | 0.936ns [0.402 – 2.179] | 0.992ns [0.743 – 1.324] | 1.413ns [0.843 – 2.369] |
| Liberia | 0.850ns [0.666 – 1.086] | 1.055ns [0.698 – 1.595] | 0.616ns [0.324 – 1.171] | 0.815ns [0.321 – 2.070] | 0.994ns [0.749 – 1.319] | 1.832*** [1.206 – 2.785] |
| Mali | 1.104ns [0.930 – 1.311] | 1.094ns [0.779 – 1.538] | 1.005ns [0.790 – 1.280] | 1.265ns [0.787 – 2.034] | 1.134ns [0.963 – 1.335] | 1.223ns [0.882 – 1.696] |
| Niger | 0.979ns [0.768 – 1.247] | 0.702ns [0.298 – 1.652] | 0.831ns [0.600 – 1.150] | 0.149* [0.016 – 1.377] | 1.028ns [0.818 – 1.291] | 0.741ns [0.338 – 1.626] |
| Nigeria | 0.932ns [0.761 – 1.142] | 0.984ns [0.722 – 1.342] | 0.953ns [0.671 – 1.352] | 0.686ns [0.365 – 1.293] | 1.163ns [0.956 – 1.414] | 1.127ns [0.828 – 1.533] |
| Senegal | 0.803ns [0.595 – 1.083] | 1.204ns [0.644 – 2.250] | 0.884ns [0.607 – 1.288] | 0.856ns [0.356 – 2.058] | 0.678*** [0.518 – 0.888] | 0.894ns [0.495 – 1.615] |
| Sierra Leone | 1.217ns [0.946 – 1.565] | 0.836ns [0.529 – 1.322] | 0.846ns [0.527 – 1.360] | 0.538ns [0.189 – 1.530] | 1.271* [0.981 – 1.646] | 0.870ns [0.535 – 1.414] |
| Togo | 1.409*** [1.120 – 1.772] | 1.537** [1.029 – 2.295] | 0.824ns [0.553 – 1.229] | 0.832ns [0.436 – 1.587] | 1.109ns [0.879 – 1.399] | 1.241ns [0.834 – 1.847] |
| Sub-Saharan Africa | 1.134*** [1.086 – 1.184] | 1.155*** [1.070 – 1.246] | 0.957ns [0.883 – 1.038] | 1.021ns [0.883 – 1.181] | 1.122*** [1.069 – 1.177] | 1.233*** [1.133 – 1.342] |
| Source: DHS – Reference modality: planned pregnancy – p value significance: *** 1%; ** 5%; * 10%; ns not significant – Logistic regressions controlled by region of residence, place of residence, occupation, media exposure and level of education of the woman, the level of education of her spouse, the age of the woman at the birth of the child, the order of birth, the interval with the previous birth, the standard of living of the household and, for supranational levels, the country of residence. | ||||||