Abstract
Objective
This study aimed to investigate the magnitude and risk factors associated with concurrent anthropometric and nutritional deficiencies among children aged 6 to 59 months in Liberia.
Materials and methods
We conducted a multilevel analysis using data from the 2019 Liberia Demographic and Health Survey (LDHS) to gain a comprehensive understanding of the problem. The study examined the prevalence of common problems, such as stunting, wasting, and anaemia, among children in Liberia to gain a deeper understanding of the issue. This study employed four different regression models. These models included panel mixed-effects Poisson regression, panel mixed-effects logistic regression, Poisson regression, and conditional fixed-effects regression.
Results
Our findings showed that, among children aged 6–59 months in Liberia, the prevalence of stunting was 31.98%. Additionally, wasting was 6.84%, underweight was 11.97%, and anaemia was 34.09%. Notably, children aged 24–35 months faced a significantly higher risk, with a 59.2% increased incidence rate of concurrent deficiencies (IRR = 1.592, p < 0.05). Female children had a 22.6% lower incidence of deficiencies compared to males (IRR = 0.774, p < 0.05). Household wealth status showed a strong protective effect, with children in the wealthiest households experiencing a 46.3% lower risk of multiple deficiencies (IRR = 0.537, p < 0.01). Complete immunisation was linked to a 37.8% reduction in the incidence of deficiencies (IRR = 0.622, p < 0.05). Furthermore, concurrent anaemia and stunting (CAS) were present in 37.0% of cases, with children in the wealthiest households having 87% lower odds (AOR = 0.130, p = 0.024).
Conclusion
By implementing effective interventions that promote proper nutrition and growth, policymakers and stakeholders can work toward improving the health and well-being of children in Liberia and reducing the burden of concurrent anthropometric and nutritional deficiencies. We therefore recommend that policymakers and practitioners in Liberia prioritize cluster-specific interventions to promote child health, given the high variability at the cluster level.
Keywords: Child malnutrition, Concurrent anthropometric deficiencies, Nutritional deficiencies, Stunting, Wasting, Anaemia, Risk factors, Multilevel analysis, Demographic and health survey, Liberia
Introduction
Child malnutrition remains a serious global health problem, especially in low- and middle-income countries where vulnerable groups face systemic obstacles to growth and development [1]. Undernutrition is responsible for nearly half of all child deaths worldwide, emphasizing its severe impact [2, 3]. Besides its immediate health effects, malnutrition has long-term impacts on a child’s physical growth, cognitive development, and overall well-being [4]. To monitor and evaluate nutritional status, standard indicators like stunting, wasting, and underweight are commonly used [5]. Recognizing the co-occurrence of anthropometric and nutritional deficiencies is essential for developing targeted interventions and effective public health policies [6]. This study seeks to examine the extent and determinants of concurrent nutritional and anthropometric deficits among children aged 6 to 59 months, addressing an important gap in knowledge and providing evidence to inform context-specific policies and programmes [2].
Liberia, situated in West Africa, faces persistent challenges that increase the risk of child malnutrition, such as widespread poverty, food insecurity, the aftermath of prolonged civil conflict, and a fragile healthcare system [7]. These socioeconomic and environmental factors create a complex landscape for child health and nutrition. The unique features of Liberia make it an important context for examining the coexistence of anthropometric and nutritional deficiencies in children. This study investigated the prevalence and causes of abnormal body parameters and nutritional deficiencies in children aged 6 to 59 months in Liberia. While stunting, wasting, and anaemia are serious conditions individually, studying their coexistence is essential because multiple nutritional issues often occur together in the same child, heightening health risks [4, 6]. Children who are both stunted and anaemic, for example, face greater threats to physical growth, cognitive development, immune function, and survival than those experiencing only one deficiency [8, 9]. Analysing these overlapping conditions helps identify the most vulnerable groups, uncover common underlying risk factors such as poverty, poor diet, and infections [5], and supports more effective integrated interventions [10]. Focusing solely on individual deficiencies might neglect the broader impact of combined health problems within affected populations.
This study provides a detailed analysis of concurrent anthropometric and nutritional deficiencies in children, contrasting with earlier research in Liberia and West Africa that primarily examined individual nutritional issues, such as stunting, wasting, or anaemia, in isolation [5, 10]. It employs robust multi-level modeling to consider both individual and community-level effects, revealing significant community variability that previous studies often overlooked. Our research provides targeted insights for developing integrated interventions to effectively combat child malnutrition in Liberia by analyzing both combined and separate deficiencies and identifying key risk factors, including socioeconomic, maternal, and environmental influences. This essential knowledge can guide policies aimed at achieving SDG targets related to hunger and health in post-conflict and economically challenged settings.
Variations in anthropometric indices can have lasting effects on a child’s physical health, cognitive development, and overall well-being [8]. It is essential to identify the core elements of this issue to develop effective interventions that can enhance children’s health and nutritional status in Liberia [10]. The unique socioeconomic and environmental characteristics of Liberia, including limited healthcare access, food insecurity, and high rates of infectious diseases, make it a vital case study for investigating child malnutrition [11]. This study aimed to assess this issue by analyzing individual, community, and regional factors.
We emphasize the negative effects of child malnutrition on health, cognitive development, physical growth, and future opportunities for Liberian children affected by prolonged civil conflict, limited healthcare, and widespread poverty. This aligns with Liberia’s commitment to the United Nations’ Sustainable Development Goals, specifically Goal 2 (Eliminating Hunger) and Goal 3 (Promoting Health and Well-being). This study aims to offer accurate, up-to-date data to support evidence-based strategies for addressing child malnutrition nationwide. This data can help governments, healthcare professionals, and humanitarian organizations allocate resources more efficiently and target interventions more effectively. The study aims to answer two main questions: What is the prevalence of concurrent anthropometric and nutritional deficiencies among children aged 6 to 59 months in Liberia? And how do individual, household, community, and regional factors influence the risk of these concurrent deficiencies in Liberian children?
Materials and methods
The Liberia Demographic and Health Survey provided information on the health and nutritional status of children in Liberia. The categorization of undernutrition in children was performed using height-for-age (HAZ), weight-for-age (WAZ), and weight-for-height (WHZ) standard deviations (SDs) from the World Health Organization (WHO), also known as z scores, to determine stunting, underweight, and wasting, respectively. Anaemia was defined as a haemoglobin level less than 11 g/dl. The study used two outcome variables originally defined by Geda, Feng, Henry, Lepnurm, Janzen, and Whiting and adopted from their 2021 study on nutritional deficiencies in Ethiopian children [12]. Multiple nutritional deficiencies were measured by a combination of wasting, underweight, stunting, anaemia, and concurrent anaemia and stunting (CAS). In the main analysis, the value of “1” was assigned if a child had any of the three anthropometric deficiencies (stunting, underweight, or wasting) or anaemia, and “0” if the child experienced none of the four nutritional problems.
For the secondary analysis, CAS was the outcome variable. For CAS, the value of “1” was used if a child was both anaemic and stunted, and “0” otherwise; thus, the outcome variable, concurrent anaemia and stunting (CAS), was constructed as a binary indicator. A child was coded as having CAS (value = 1) if they were both stunted (height-for-age z-score below − 2 SD) and anaemic (haemoglobin level < 11 g/dL), and as not having CAS (value = 0) if they had neither or only one of these conditions. This combined indicator captures children experiencing both chronic undernutrition and micronutrient deficiency simultaneously, reflecting a higher-risk subgroup with potentially compounded health consequences. The key study variables and their operational definitions are summarized in Table 1 (see appendix).
The selection of the explanatory variables for this research was based on the following characteristics: child-level variables (child age in months, sex of child, preceding birth interval, birth order number), household-level variables (respondents’ current age, religion, wealth index, highest educational level, currently breastfeeding), childcare practices (duration of breastfeeding), and community-level variables (native language of respondent) that were used in other studies, along with the availability of variables in the dataset and statistical validity. We included additional explanatory variables such as entries in the immunization roster, type of toilet, time to reach the water source, whether the child was given plain water, respondent occupation, vitamin A supplementation in the last six months, and shared toilet facilities in our model.
We analyzed the data using Stata 17. Analyses were weighted to account for sampling probabilities, and the stratification and clustering of the data were taken into consideration. Descriptive analysis was conducted to examine the characteristics of the study sample. All descriptive analyses and single-level regression models in this study were weighted using the sampling weights provided by the 2019 Liberia Demographic and Health Survey (LDHS) to ensure national representativeness. However, for the multilevel (mixed effects) models, weights were not directly applied because standard multilevel modeling procedures in Stata do not easily accommodate complex survey weights at multiple levels. As a result, multilevel analyses were performed without weighting, and the findings should be interpreted with this limitation in mind.
We used four regression models, including Poisson regression, mixed-effects Poisson regression, conditional fixed-effects regression, and mixed-effects logistic regression, to thoroughly examine the factors influencing concurrent anthropometric and nutritional deficiencies. The Poisson regression model is well-suited for analyzing rare event data and provides consistent estimates under the assumption of equi-dispersion [13]. Due to the hierarchical structure of the dataset, where children are nested within families and communities, the mixed-effects Poisson regression incorporates random effects to address unobserved heterogeneity and intra-cluster correlations [14]. Moreover, the conditional fixed-effects regression reduces unobserved time-invariant confounding within clusters by focusing on within-cluster variance [15]. Finally, the mixed-effects logistic regression is used for the binary outcome of concurrent anaemia and stunting (CAS), allowing the model to incorporate both fixed and random effects, thereby improving the estimation of associations in multilevel data [16]. We employed multiple statistical models to capture different aspects of the data, account for clustering and unobserved differences within communities, and provide robust, reliable estimates of parameters related to nutritional deficiencies. These models enhance our analytical approach by addressing both fixed and random sources of variation, which aligns with the multilevel nature of the data.
Ethical clearance
Demographic and Health Surveys (DHSs) have been conducted in many countries. DHS is a program that collects and shares nationally representative data on key demographic and health indicators. To access the dataset, we first applied to the Demographic and Health Surveys (DHS). It took about two months to receive the data after submission. The dataset was provided in response to our application. During the process, we had to outline relevant information and justify the research purpose. DHS verified that the dataset met ethical and privacy standards before granting access. This dataset includes comprehensive socioeconomic, health, and demographic data. We did not perform direct data collection from participants. However, ethical standards were maintained by using secondary data from the 2019 Liberia Demographic and Health Survey (LDHS). The LDHS data collection received prior ethical approval from the appropriate institutional review boards, and informed consent was obtained from all participants during the initial survey. The dataset shared with researchers was fully anonymized, ensuring confidentiality and privacy for all individuals.
Results
Table 4 (see appendix) shows the proportion of children aged 6–59 months with stunting, wasting, underweight, or anaemia based on demographic characteristics. Overall, malnutrition remains a major public health issue among Liberian children in this age group. Nearly one-third are stunted (31.98%), and more than a third are affected by anaemia (34.09%). Although wasting is less common overall (6.84%), specific subgroups face higher risks. For example, children aged 24–35 months showed a statistically significant increase in the incidence of multiple deficiencies (IRR = 1.59, p < 0.05). Although the data show a 50.00% prevalence of wasting among children aged 24–35 months, this figure is based on a very small sample size (n = 2). Therefore, it should not be regarded as a reliable indicator of vulnerability in this age group. Caution is recommended when interpreting this finding, as the limited sample size affects both statistical reliability and the broader applicability of the result.
Table 4.
Proportion of children (6–59 months) with stunting, wasting, underweight or anaemia from demographic characteristics
| Characteristic | Stunting N (%) |
Wasting N (%) |
Underweight N (%) |
Anaemia N (%) |
|---|---|---|---|---|
| Child Level | ||||
| Child’s age in months | ||||
| 6–23 months | 134 (26.48) | 35 (6.89) | 75 (14.71) | 386 (73.38) |
| 24-35months | 2 (50.00) | 0 (0.00) | 1 (25.00) | 4 (100.00) |
| Sex of the child | ||||
| male | 402 (33.53) | 18 (7.09) | 39 (15.23) | 198 (34.80) |
| female | 380 (30.50) | 17 (6.59) | 37 (14.34) | 192 (33.39) |
| Preceding birth interval (months) | ||||
| 6–11 months | 2 (100.00) | 0 (0.00) | 2 (100.00) | 1 (20.00) |
| 12–23 months | 17 (36.96) | 6 (13.04) | 9 (19.15) | 42 (40.38) |
| 24 + months | 94 (24.42) | 27 (6.98) | 55 (14.18) | 286 (33.33) |
| birth order number | ||||
| first | 23 (30.26) | 2 (2.63) | 10 (13.16) | 60 (34.68) |
| second and above | 113 (26.04) | 33 (7.57) | 66 (15.07) | 330 (33.99) |
| Vitamin A in the last 6 months | ||||
| no | 68 (25.00) | 24 (8.82) | 44 (16.06) | 204 (31.29) |
| yes | 67 (28.51) | 11 (4.64) | 32 (13.50) | 183 (37.58) |
| do not know | 1 (33.33) | 0 (0.00) | 0 (0.00) | 3 (60.00) |
| Household Level | ||||
| Respondent’s current age | ||||
| 15–25 years | 55 (29.41) | 6 (3.68) | 26 (13.90) | 125 (35.51) |
| 26–35 years | 49 (23.11) | 16 (7.92) | 30 (14.08) | 163 (33.33) |
| 36 years and above | 32 (28.83) | 13 (8.84) | 20 (17.54) | 102 (33.66) |
| Religion | ||||
| Orthodox | 26 (33.33) | 8 (10.13) | 12 (15.19) | 59 (29.50) |
| Muslim | 110 (25.46) | 27 (6.24) | 64 (14.71) | 331 (35.06) |
| Wealth index combined | ||||
| poorest | 297 (12.15) | 11 (6.43) | 30 (17.54) | 127 (31.91) |
| poorer | 227 (33.38) | 10 (6.76) | 24 (16.11) | 109 (33.44) |
| middle | 152 (31.93) | 8 (7.92) | 14 (13.73) | 88 (41.71) |
| richer | 79 (28.01) | 5 (9.43) | 5 (9.43) | 36 (28.57) |
| richest | 27 (15.88) | 1 (2.56) | 3 (7.69) | 30 (36.14) |
| Highest educational level | ||||
| no education | 341 (32.85) | 16 (6.87) | 41 (17.60) | 171 (32.95) |
| primary level | 254 (34.23) | 13 (8.90) | 22 (14.86) | 117 (35.24) |
| secondary and above | 187 (28.12) | 6 (4.51) | 13 (9.77) | 102 (34.81) |
| Currently breastfeeding | ||||
| no | 352 (33.72) | 6 (7.14) | 9 (10.71) | 64 (37.65) |
| yes | 430 (30.69) | 29 (6.78) | 67 (15.58) | 326 (33.47) |
| Respondent’s occupation | ||||
| not working | 205 (29.16) | 10 (6.99) | 10 (6.99) | 114 (33.33) |
| agriculture sector | 137 (31.93) | 6 (6.06) | 6 (6.06) | 81 (41.12) |
| other sectors | 437 (33.36) | 19 (7.06) | 19 (7.06) | 194 (32.17) |
| Childcare practices | ||||
| Duration of breastfeeding | ||||
| ever breastfed, not currently breastfeeding | 500 (36.15) | 6 (7.41) | 160 (11.52) | 62 (38.51) |
| never breastfed | 17 (45.95) | 0 (0.00) | 4 (10.53) | 2 (22.22) |
| still breastfeeding | 265 (25.85) | 29 (6.78) | 131 (12.63) | 326 (33.47) |
| Toilet facilities shared with other households | ||||
| no | 140 (29.60) | 7 (7.95) | 9 (10.00) | 303 (28.56) |
| yes | 234 (30.27) | 11 (6.88) | 24 (15.00) | 477 (26.30) |
| not a dejure resident | 15 (39.47) | 1 (16.67) | 2 (33.33) | 20 (15.04) |
| Entries in immunization roster | ||||
| partially immunized | 543 (31.90) | 5 (12.50) | 208 (12.14) | 356 (34.87) |
| fully immunized | 92 (36.36) | 0 (0.00) | 44 (17.05) | 33 (27.97) |
| Gave the Child Plain Water | ||||
| no | 91 (27.91) | 2 (8.00) | 28 (8.48) | 20 (12.82) |
| yes | 364 (31.24) | 33 (6.78) | 161 (13.68) | 370 (37.45) |
| Community Level | ||||
| Native language of the respondent | ||||
| English | 782 (31.98) | 35 (6.84) | 295 (11.97) | 390 (34.09) |
| Type of place of residence | ||||
| urban | 228 (28.79) | 11 (7.24) | 63 (10.18) | 119 (33.90) |
| rural | 554 (33.51) | 24 (6.67) | 213 (12.78) | 271 (34.17) |
| Toilet type | ||||
| others | 389 (30.30) | 68 (5.28) | 158 (12.21) | 199 (34.79) |
| no facility/bush/field | 393 (33.85) | 37 (3.17) | 137 (11.71) | 191 (33.39) |
| Time to get to the water source | ||||
| less than 30 min | 631 (33.07) | 90 (4.70) | 229 (11.92) | 1178 (26.74) |
| 31–60 min | 59 (28.50) | 6 (2.88) | 25 (12.02) | 133 (29.49) |
| 1 h and above | 92 (27.88) | 9 (2.69) | 41 (12.24) | 221 (26.06) |
Source: Authors’ construct (2023)
Moreover, male children tend to have higher rates of stunting and underweight, while females show slightly higher rates of anaemia. Studies suggest that male children may have higher nutritional needs and different biological susceptibilities, which could potentially explain these sex differences [17]. Socioeconomic factors play a key role, as children from the poorest households are more likely to be stunted and anaemic. Conversely, children in the richest quintile have an 87% lower chance of suffering from both anaemia and stunting (AOR = 0.13, p < 0.05). Full immunization acts as a protective factor, reducing the occurrence of multiple nutritional deficiencies by nearly 38% (IRR = 0.62, p < 0.05). These findings emphasize the need for targeted, equity-focused strategies that address both individual and combined nutritional deficiencies.
Table 2, in the appendix, presents the results of the chi-square analysis examining the relationships between explanatory variables and the simultaneous presence of stunting and anaemia. According to Table 2, the study identified strong associations between specific characteristics and the co-occurrence of these conditions in children aged 6–59 months in Liberia. Notably, receiving vitamin A within the past 6 months emerged as a significant factor (p < 0.05). This highlights the potential of targeted nutritional interventions to reduce the combined occurrence of stunting and anaemia, underscoring the ongoing importance of comprehensive vitamin supplementation programs to improve child health. As shown in Table 2, among childcare practices, including names on the immunization list, was found to be a very important factor related to both stunting and anaemia occurring together (p < 0.001). This finding emphasizes the critical role of organized immunization programs in decreasing the prevalence of these health issues among children in Liberia.
Table 2.
Results of chi-square analysis for associations between the explanatory variables and concurrent stunting and anaemia
| Characteristics | No CAS N (%) |
CAS N (%) |
Chi-square |
|---|---|---|---|
| Child Level | 0.96 | ||
| Child’s age in months | |||
| 6–23 months | 102 (19.39) | 424 (80.61) | |
| 24-35months | 0 (0.00) | 4 (100.00) | |
| Sex of the child | |||
| male | 353 (62.04) | 216 (37.96) | |
| female | 363 (63.13) | 212 (36.87) | |
| Preceding birth interval in months | |||
| 6–11 months | 3 (60.00) | 2 (40.00) | 1.15 |
| 12–23 months | 60 (57.69) | 44 (42.31) | |
| 24 + months | 541 (63.05) | 317 (36.95) | |
| Birth order number | |||
| first | 109 (63.01) | 64 (36.99) | 0.02 |
| second and above | 607 (62.51) | 364 (37.49) | |
| Vitamin A in the last 6 months | |||
| no | 428 (65.64) | 224 (34.36) | 6.79** |
| yes | 286 (58.73) | 201 (41.27) | |
| do not know | 2 (40.00) | 3 (60.00) | |
| Household Level | |||
| Respondents’ current age | |||
| 15–25 years | 258 (62.17) | 157 (37.83) | 0.13 |
| 26–35 years | 309 (63.19) | 180 (36.81) | |
| 36 years and above | 149 (62.08) | 91 (37.92) | |
| Religion | |||
| Orthodox | 132 (66.00) | 68 (34.00) | 1.21 |
| Muslim | 584 (61.86) | 360 (38.14) | |
| Wealth index combined | |||
| poorest | 256 (64.32) | 142 (35.68) | 6.89 |
| poorer | 202 (61.96) | 124 (38.04) | |
| middle | 119 (56.40) | 92 (43.60) | |
| richer | 88 (69.84) | 38 (69.84) | |
| richest | 51 (61.45) | 32 (38.55) | |
| Highest level of education | |||
| no education | 327 (63.01) | 192 (36.99) | 1.98 |
| primary | 205 (61.75) | 127 (38.25) | |
| secondary | 170 (61.82) | 105 (38.18) | |
| higher | 14 (77.78) | 4 (22.22) | |
| Currently breastfeeding | |||
| never | 98 (57.65) | 72 (42.35) | 2.08 |
| breastfeed | 618 (63.45) | 356 (36.55) | |
| Respondent’s Occupation | |||
| not working | 219 (64.04) | 123 (35.96) | 3.99 |
| agriculture sector | 111 (56.35) | 86 (43.65) | |
| other sectors | 385 (63.85) | 218 (36.15) | |
| Child Care Practices | |||
| Duration of breastfeeding | |||
| ever breastfed, not currently breastfeeding | 92 (57.14) | 69 (42.86) | 2.41 |
| never breastfed | 6 (66.67) | 3 (33.33) | |
| still breastfeeding | 618 (63.45) | 356 (36.55) | |
| Toilet facilities shared with other households | |||
| no | 113 (60.11) | 75 (39.89) | 1.01 |
| yes | 237 (64.05) | 133 (35.95) | |
| not a dejure resident | 8 (57.14) | 6 (42.86) | |
| Entries in the immunization roster | |||
| partially immunized | 2518 (61.80) | 1262 (38.20) | 28.11*** |
| fully immunized | 513 (68.64) | 175 (31.36) | |
| additional vaccinations | 29 (80.00) | 6 (20.00) | |
| Gave the child plain water | |||
| no | 542 (86.54) | 148 (13.46) | 58.33*** |
| yes | 1543 (58.81) | 906 (41.19) | |
| Community Level | |||
| Native language of the respondent | |||
| English | 3,925 (68.81) | 1,779 (31.19) | |
| Type of place of residence | |||
| urban | 1,335 (70.49) | 559 (29.51) | 3.70** |
| rural | 2590 (67.98) | 1220 (32.02) | |
| Toilet type | |||
| others | 2086 (69.35) | 922 (30.65) | 0.86 |
| no facility/bush/field | 1839 (68.21) | 857 (31.79) | |
| Time to get to the water source | |||
| less than 30 min | 3022 (68.60) | 1383 (31.40) | 2.87 |
| 31–60 min | 301 (66.74) | 150 (33.26) | |
| 1 h and above | 602 (70.99) | 246 (29.01) | |
*** p < 0.0, ** p < 0.05, * p < 0.1.
Source: Authors’ construct (2023)
Effective immunization programs are crucial for enhancing children’s overall health and well-being. Additionally, providing plain water for children is critically important, as it is associated with both stunting and anaemia co-occurring (p < 0.001). This finding highlights the need for further investigation into water-related practices, sanitation, and hygiene standards. These factors should be considered when developing targeted interventions to improve water quality and hygiene, which may influence children’s health. However, further research is needed to confirm the direction and magnitude of this relationship. Concerning community-level factors, the specific type of residential area showed a marginally significant relationship (p < 0.10) with the simultaneous occurrence of stunting and anaemia.
The findings of the multilevel regression analysis, as shown in Table 3 (appendix), demonstrate the linkage between concurrent anthropometric and nutritional deficiencies in children aged 6–59 months in Liberia. The research employed Poisson regression, conditional fixed-effects regression, mixed-effects Poisson regression, and mixed-effects logistic regression to systematically identify consistent predictors across various model parameters. Essential findings indicate substantial demographic, social, and health-related factors that interact to affect child nutritional outcomes. The random effects at the cluster level were highly significant (p < 0.001), confirming that unobserved community-level factors have a substantial impact on dietary outcomes. This validates the application of multilevel models and suggests the potential value of decentralized, context-specific interventions to address unobserved community-level influences on nutrition [10].
Table 3.
Estimated regression models showing factors associated with nutritional deficiencies
| Characteristic | Model 1 AOR |
Model 2 AOR (p-value) |
Model 3 IRR |
Model 4 (CAS) AOR |
|---|---|---|---|---|
| Random Effect | ||||
| cluster | 6.51e-12*** | 0.351 | ||
| Fixed Effect | ||||
| Age of the child | ||||
| 24-35months | 1.592** | 1.330 | ||
| Sex of the child | ||||
| female | 0.774** | 0.984*** | 0.877 | 0.628 |
| Preceding birth interval | ||||
| 12–23 months | 0.91 | 1.24 | 1.42 | |
| 24 + months | 0.57 | 0.43*** | 0.93 | |
| Vitamin A in the last 6 months | ||||
| yes | 0.91 | 0.75*** | 0.93 | 1.84 |
| do not know | 1.31 | 1.24 | ||
| Mother’s age | ||||
| 26–35 years | 1.09 | 1.27*** | 1.17 | 0.91 |
| 36 years and above | 1.35* | 1.99*** | 1.38 | 0.54 |
| Religion | ||||
| Muslim | 0.78* | 0.24*** | 0.87 | 0.86 |
| Wealth index | ||||
| poorer | 0.92 | 1.29*** | 1.03 | 1.06 |
| middle | 0.87 | 1.51*** | 0.98 | 1.01 |
| richer | 0.54** | 0.82*** | 0.68 | 0.13** |
| richest | 0.69 | 0.46*** | 0.73 | 0.26 |
| Mother’s education | ||||
| primary level | 1.13 | 1.32*** | 1.17 | 0.94 |
| secondary and above | 1.06 | 1.04*** | 0.98 | 1.34 |
| Currently breastfeeding | ||||
| breastfeed | 1.03 | 0.98*** | 0.89 | 0.40 |
| Occupation | ||||
| agriculture sector | 1.26 | 1.93*** | 0.93 | 1.76 |
| other sectors | 1.05 | 1.13*** | 0.90 | 0.68 |
| Toilet facilities are shared with others | ||||
| yes | 0.86 | 1.10*** | 1.08 | 1.05 |
| not a dejure | 1.47 | 0.29*** | 1.81 | |
| Immunization | ||||
| fully immunized | 0.62** | 0.88*** | 0.89 | 1.16 |
| Gave the child plain water | ||||
| yes | 0.95 | 0.99*** | 1.03 | 0.44 |
| Residence | ||||
| rural | 0.96 | 0.92 | 0.31* | |
| Time to get to the water source | ||||
| 31–60 min | 0.97 | 1.09*** | 1.21 | 0.87 |
| 1 h and above | 0.86 | 1.35*** | 1.00 | 0.82 |
| Constant | 0.00*** | 1.65 | 95.69*** | |
*** p < 0.01, ** p < 0.05, * p < 0.1. All analyses for conditional fixed-effects regression were weighted. Model 1 is the Poisson regression, Model 2 is the Conditional Fixed-Effect regressio, Model 3 is the Panel Mixed Effects Poisson regression and Model 4 is the Mixed Effect Logistic regression. AOR Adjusted Odds Ratio, IRR Incidence Rate Ratio
Source: Authors’ construct (2023)
Child-level factors were found to be important. Age consistently predicted dietary deficiencies, especially in children aged 24–35 months, who were much more prone to multiple deficiencies. This age coincides with the transition from breastfeeding to complementary foods, a period often associated with inadequate nutrient intake and increased susceptibility to infections [18]. The Poisson regression model showed a 59.2% increase in incidence (IRR = 1.59, p < 0.05), while the conditional fixed-effects model indicated an increased risk (AOR = 1.33), although it was not statistically significant. This sensitivity is related to dietary changes during the post-weaning phase. It aligns with earlier research, which highlights that the ages of 24–35 months are a critical stage for child growth and nutritional interventions [6, 12]. Sex differences were also significant across the models. Female children had a 22.6% lower incidence of deficiencies compared to males (IRR = 0.77, p < 0.05, using a Poisson model). They were less likely to experience concurrent stunting and anaemia in the fixed-effects model (AOR = 0.98, p < 0.001). This finding supports evidence suggesting a biological and behavioral advantage for females in early infancy nutritional outcomes [17].
The conditional fixed-effects model demonstrated that children born following a birth interval of at least 24 months exhibited significantly improved nutritional outcomes (AOR = 0.43, p < 0.001). This finding aligns with [4], suggesting a possible role of adequate birth spacing in improving maternal caregiving capacity and child nutritional outcomes. This variable was not significant in the Poisson model, indicating that fixed-effects models may better capture within-cluster differences. Health interventions such as vitamin A supplementation and vaccination showed varied yet substantial effects. Children who had received vitamin A in the previous six months had lower odds of deficiencies in the fixed-effects model (AOR = 0.75, p < 0.001). However, this association was not significant in the Poisson regression. Immunization status was one of the most impactful protective factors. Fully immunized children experienced a 37.8% decrease in deficiency frequency (IRR = 0.62, p = 0.03, Poisson model) and a 12.4% reduction in odds in the mixed-effects model (AOR = 0.88, p < 0.001). These results support other studies linking vaccination coverage with improved nutritional outcomes [19, 20].
Maternal traits had a substantial influence. Children of mothers aged 36 and older showed increased nutritional sensitivity, with a 34.5% higher incidence (IRR = 1.35, p < 0.10) in the Poisson model and a 99.2% increase in odds in the fixed-effects model (AOR = 1.99, p < 0.001). This may reflect possible biological or socioeconomic disadvantages among older mothers, a trend previously noted in the literature [21]. Maternal education offered inconsistent protective benefits across models, indicating that knowledge alone might be insufficient without proper access to health and nutrition services [22]. Religion became a significant factor. In the Poisson model, children from Muslim households experienced a 22.4% reduction in deficiency incidence (IRR = 0.78, p < 0.10), while the fixed-effects model showed a markedly lower odds ratio (AOR = 0.24, p < 0.001). These results may be influenced by unmeasured factors, including cultural practices or dietary habits, which require further qualitative research to clarify.
Socioeconomic status regularly offers protection. Children from affluent families showed a 46.3% lower incidence rate in the Poisson model (IRR = 0.54, p < 0.05) and a 74% reduced likelihood of simultaneous anaemia and stunting in the mixed-effects logistic model (AOR = 0.13, p < 0.05). These results point out the robust relationship between household wealth and child health outcomes, primarily due to improved access to nutrition, healthcare, and sanitation [11]. Environmental and sanitation factors showed less consistent but still significant associations. Breastfeeding was unexpectedly associated with increased odds of deficiencies in the fixed-effects model, which may reflect reverse causality or unmeasured confounding, such as delays in complementary feeding. Access to water and sanitation variables, such as the time required to reach a water source and the availability of shared toilet facilities, showed inconsistent effects, necessitating further investigation into hygiene-related factors that impact child malnutrition.
Discussion of findings
Our findings align with those of [9], who reported a 72.6% prevalence of anaemia among Southeast Asian children aged 6–36 months, with key predictors including young age, maternal anaemia, low height-for-age, low family income, maternal education, unboiled water intake, and recent fever. To combat regional anaemia and malnutrition, urgent comprehensive actions were recommended. Similarly [20], found a 21.4% prevalence of vitamin A deficiency among Brazilian children aged 6–24 months, noting that extended breastfeeding was protective while low birth weight increased vulnerability [17]. also observed significant sex-based differences in dietary diversity among Indian adolescents, with boys consuming more protein- and vitamin-rich foods than girls at age 15. To address sex-based nutritional gaps, the study suggested targeted interventions. The N@C programme in Bangladesh, as evaluated by [23], was shown to improve maternal and child nutrition indicators, including prenatal care uptake and dietary diversity. Local planning and multi-sectoral approaches contributed to these improvements.
[24] linked child undernutrition in Tanzanian fishing communities to low birth weight, inadequate immunisation, limited maternal education, and poor access to health services. We recommend enhancing healthcare and empowering mothers through education. In Ghana [21], reported a 64.3% vitamin A supplementation rate but highlighted a lack of nutritional awareness among caregivers as a significant barrier to uptake. The study emphasized the importance of caregiver nutrition education to increase vitamin A supplementation [19]. examined vaccination status and child nutrition in 16 countries. Vaccination is fundamental for child health, and incomplete immunization is associated with underweight, wasting, and stunting. On Leti Island, Indonesia [25], found that exclusive breastfeeding and improved healthcare access significantly reduced childhood stunting. The study advised increasing healthcare access and regular health visits for mothers and children [22]. demonstrated that maternal education of more than 10 years was associated with reduced rates of stunting, underweight, and wasting in Malawi, Tanzania, and Zimbabwe. The study highlighted the importance of maternal education in infant nutrition [26]. showed that the American College of Cardiology/American Heart Association (ACC/AHA) criteria identified a higher incidence of hypertension among young adults in Albania compared to the Report of the Joint National Committee (JNC7) guidelines. Age, education, wealth, number of living children, health insurance status, and sex were associated with hypertension.
Our findings align with previous studies across sub-Saharan Africa, where similarly high rates of malnutrition have been reported, such as the 72.6% anaemia prevalence among Southeast Asian children [9] and high rates of concurrent stunting and wasting in Ethiopia and Uganda [4, 6]. However, our observed stunting prevalence of 31.98% and anaemia prevalence of 34.09% in Liberia are somewhat lower than the figures reported by [10] during post-conflict periods. The increased vulnerability among children aged 24–35 months may result from biological factors, including weaning from breastfeeding and increased exposure to infections at this stage, along with changing dietary practices that may not meet nutritional needs [8]. Additionally, male children often experience higher rates of malnutrition, possibly due to higher physiological demands for growth, different care practices, or cultural feeding preferences that might favor female children in some contexts [17]. By placing our results within this broader context, we highlight the persistent and complex nature of child undernutrition in Liberia, emphasizing the need for age- and sex-specific interventions.
The findings of significant cluster-level variation and specific risk factors for concurrent nutritional deficiencies in Liberian children have important policy implications. They suggest that policymakers should develop interventions tailored to each community’s unique characteristics instead of implementing uniform national policies. For example, cluster-specific interventions could include targeted nutrition education for mothers in high-risk areas, localized supplementation programs, improved access to health services, and community-led efforts to strengthen water and sanitation infrastructure. However, implementing these tailored approaches faces barriers, including limited financial resources, shortages of trained health personnel, logistical challenges in reaching remote communities, and potential political and administrative fragmentation that can hinder coordination across regions. Additionally, collecting and maintaining detailed local data necessary for designing and monitoring cluster-specific interventions can overburden already fragile health systems. Overcoming these barriers will require ongoing investment, capacity-building, and close collaboration between government agencies, local leaders, and international partners to ensure that interventions are both culturally appropriate and sustainable.
This study has several limitations. First, it relies on cross-sectional data from the 2019 Liberia Demographic and Health Survey, which limits causal inference. Second, although multilevel modelling enhances estimation by addressing clustering, residual confounding may still exist due to unmeasured variables such as dietary intake or maternal mental health. Third, certain subgroups in the dataset (e.g., children aged 24–35 months in some analyses) had very small sample sizes, which restricts statistical power and the reliability of estimates in these strata. Fourth, anaemia was measured solely using haemoglobin levels without specifying type or cause, thereby limiting interpretability. Lastly, while this study considers multiple nutritional outcomes, the exclusion of other micronutrient deficiencies or environmental exposures may underestimate the true burden of malnutrition.
Conclusion and recommendations
This study revealed that many Liberian children aged 6–59 months face both anthropometric and nutrient deficiencies, with nearly one-third experiencing stunting and over one-third affected by anaemia. Children aged 24 to 35 months and males were identified as the most vulnerable groups, with socioeconomic factors such as household wealth and maternal education playing a significant role in nutritional outcomes. The results also showed notable variation at the cluster level, indicating that unobserved community-specific factors have a significant influence on nutritional health.
Based on these findings, we recommend that policymakers consider developing and testing cluster-specific interventions tailored to the varied needs of local communities. These could include targeted nutrition education, enhanced immunisation coverage, and upgraded water and sanitation infrastructure, especially for high-risk age groups and male children. Evidence-based interventions, including micronutrient supplementation, community health programs, and improved feeding practices, have shown promise and cost-effectiveness in similar contexts [18, 27]. However, successful implementation will require addressing challenges such as resource limitations, logistical issues, and the need for local data to inform actions. Continued investment in data systems, community engagement, and integrated health and nutrition strategies will be crucial to effectively reducing child malnutrition in Liberia. Future studies should employ longitudinal designs to investigate the causal relationships between identified risk factors and nutritional outcomes, as well as to examine how seasonal changes, infections, and dietary patterns contribute to concurrent deficiencies. Additionally, qualitative research could provide a deeper understanding of the cultural, behavioral, and community-specific factors affecting child nutrition in Liberia.
Acknowledgements
We acknowledge the 2019 Liberia Demographic and Health Survey (LDHS) for provision of the data for this study.
Abbreviations
- DHS
Demographic and Health Survey
- LDHS
Liberia Demographic and Health Survey
- HAZ
Height–for–Age Z–score
- WAZ
Weight–for–Age Z–score
- WHZ
Weight–for–Height Z–score
- CAS
Concurrent Anaemia and Stunting
- IRR
Incidence Rate Ratio
- AOR
Adjusted Odds Ratio
- SD
Standard Deviation
- WHO
World Health Organization
- SDGs
Sustainable Development Goals
Appendix
Table 1.
Operational definitions of key variables
| Variable | Operational Definition |
|---|---|
| stunting | Height-for-age z-score below − 2 SD (WHO standard). |
| wasting | Weight-for-height z-score below − 2 SD (WHO standard). |
| underweight | Weight-for-age z-score below − 2 SD (WHO standard). |
| anaemia | The haemoglobin level is less than 11 g/dL. |
| CAS (concurrent anaemia and stunting) | Child coded as “1” if both stunted and anaemic, “0” otherwise. |
| wealth index | Composite measures from household assets, housing materials, and access to water/sanitation, categorized into quintiles (poorest, poorer, middle, richer, richest). |
| maternal education | Highest schooling attained, classified as no education, primary, secondary, or higher. |
| currently breastfeeding | Whether the child was still receiving breast milk at the time of the survey. |
| vitamin A supplementation | Whether the child received a vitamin A dose in the past six months. |
| shared toilet facilities | Sanitation facilities shared with other households (yes/no). |
| place of residence | Urban or rural, based on DHS classification. |
| time to water source | Minutes reported to reach the primary water source. |
| child’s age and sex | Recorded in months; sex as male or female. |
| birth order | Sequence number of a child among siblings. |
| immunization status | Classified as partially, fully immunized, or additional vaccinations recorded. |
Source: Authors’ construct (2023)
Author contributions
E.G.G. - conceived the research idea, writing- original draft, data analysis, review, final approval and agreeing to the accuracy of the work. I.A. - data request, writing- original draft, data analysis, formatting of tables, review, final approval, and agreeing to the accuracy of the work. J.K.T. - conceptualisation, writing- original draft, data analysis, formatting of tables, compilation of paper, review, final proofreading and agreeing to the accuracy of the work.
Funding
No funding.
Data availability
The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
The data used in this study were obtained from the 2019 Liberia Demographic and Health Survey (LDHS), with permission for data access granted through an online request submitted to http://www.dhsprogram.com. The researchers did not interact with respondents; hence, there was no need for ethical clearance. The data used were publicly available with no personal identifiers.
Consent for publication
Not Applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Sabbahi M, Li J, Davis C, Downs SM. The role of the sustainable development goals to reduce the global burden of malnutrition. In Advances in food security and sustainability. Elsevier. 2018;(3):277–333
- 2.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. Lancet. 2013;382(9890):427–51. [DOI] [PubMed] [Google Scholar]
- 3.UNICEF. The State of the World’s Children. 2019: Children, Food and Nutrition [Internet]. New York: UNICEF; 2019 [cited 2025 Jun 27]. Available from: https://www.unicef.org/reports/state-of-worlds-children-2019
- 4.Odei Obeng-Amoako GA, Karamagi CA, Nangendo J, Okiring J, Kiirya Y, Aryeetey R, et al. Factors associated with concurrent wasting and stunting among children 6–59 months in Karamoja. Uganda Maternal Child Nutr. 2021;17(1):e13074. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kumeh OW, Fallah MP, Desai IK, Gilbert HN, Silverstein JB, Beste S, et al. Literacy is power: structural drivers of child malnutrition in rural Liberia. BMJ Nutr Prev Health. 2020;3(2):295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Roba AA, Assefa N, Dessie Y, Tolera A, Teji K, Elena H, et al. Prevalence and determinants of concurrent wasting and stunting and other indicators of malnutrition among children 6–59 months old in kersa, Ethiopia. Matern Child Nutr. 2021;17(3):e13172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Victora CG, Adair L, Fall C, Hallal PC, Martorell R, Richter L, et al. Maternal and child undernutrition: consequences for adult health and human capital. Lancet. 2008;371(9609):340–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Branca F, Ferrari M. Impact of micronutrient deficiencies on growth: the stunting syndrome. Annals Nutr Metabolism. 2002;46(Suppl 1):8–17. [DOI] [PubMed] [Google Scholar]
- 9.Zhao A, Zhang Y, Peng Y, Li J, Yang T, Liu Z, et al. Prevalence of anaemia and its risk factors among children 6–36 months old in Burma. Am J Trop Med Hyg. 2012;87(2):306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Brault MA, Kennedy SB, Haley CA, Clarke AT, Duworko MC, Habimana P, et al. Factors influencing rapid progress in child health in post-conflict liberia: a mixed methods country case study on progress in child survival, 2000–2013. BMJ Open. 2018;8(10):e021879. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Pawlak K, Kołodziejczak M. The role of agriculture in ensuring food security in developing countries: considerations in the context of the problem of sustainable food production. Sustainability. 2020;12(13):5488. [Google Scholar]
- 12.Geda NR, Feng CX, Henry CJ, Lepnurm R, Janzen B, Whiting SJ. Multiple anthropometric and nutritional deficiencies in young children in ethiopia: a multi-level analysis based on nationally representative data. BMC Pediatr. 2021;21:1–2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Scott Long J, Freese J. Regression models for categorical dependent variables using Stata. College Station, TX: StataCorp LP; 2006. [Google Scholar]
- 14.Rabe-Hesketh S, Skrondal A. Multilevel and longitudinal modeling using Stata. STATA. 2008
- 15.Allison PD. Fixed effects regression models. SAGE. 2009
- 16.Goldstein H. Multilevel statistical models. Wiley. 2011
- 17.Aurino E. Do boys eat better than girls in india?? Longitudinal evidence on dietary diversity and food consumption disparities among children and adolescents. Econ Hum Biology. 2017;25:99–111. [DOI] [PubMed] [Google Scholar]
- 18.Lutter CK, Daelmans BM, de Onis M, Kothari MT, Ruel MT, Arimond M, et al. Undernutrition, poor feeding practices, and low coverage of key nutrition interventions. Pediatrics. 2011;128(6):e1418–27. [DOI] [PubMed] [Google Scholar]
- 19.Solis-Soto MT, Paudel D, Nicoli F. Relationship between vaccination and nutritional status in children. Demographic Res. 2020;42:1–4. [Google Scholar]
- 20.Ferraz IS, Daneluzzi JC, Vannucchi HJ. Vitamin A deficiency in children aged 6 to 24 months in São Paulo state, Brazil. Nutr Res. 2000;20(6):757–68. [Google Scholar]
- 21.Hadzi D, Asalu GA, Avedzi HM, Appiah PK, Tarkang EE. Vitamin A supplementation coverage and correlates of uptake among children 6–59 months in the South Dayi district, Ghana. Cent Afr J Public Health. 2016;2(2):89–98. [Google Scholar]
- 22.Makoka D, Masibo PK. Is there a threshold level of maternal education sufficient to reduce child undernutrition? Evidence from malawi, Tanzania and Zimbabwe. BMC Pediatr. 2015;15:1–0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Nguyen PH, Sununtnasuk C, Christopher A, Ash D, Ireen S, Kabir R, et al. Strengthening nutrition interventions during antenatal care improved maternal dietary diversity and child feeding practices in urban bangladesh: results of a quasi-experimental evaluation study. J Nutr. 2023;153(10):3068–82. [DOI] [PubMed] [Google Scholar]
- 24.Mtoi EH, Nyaruhucha CN. Childcare practices and nutritional status of under-five children in tanzania: evidence from fishing communities in Pangani district. Int J Asian Social Sci. 2019;9(7):390–405. [Google Scholar]
- 25.Unmehopa A, Palutturi S, Arifin MA, Thamrin Y, Mallongi A. Accessibility of mother and child health services to stunting incidence in Leti island, Southwest Maluku district. Pharmacognosy J. 2023;15(5):856–860.
- 26.Ahammed B, Maniruzzaman M, Talukder A, Ferdausi F. Prevalence and risk factors of hypertension among young adults in Albania. High Blood Press Cardiovasc Prev. 2021;28:35–48. [DOI] [PubMed] [Google Scholar]
- 27.Bhutta ZA, Das JK, Rizvi A, Gaffey MF, Walker N, Horton S, et al. Evidence-based interventions for improvement of maternal and child nutrition: what can be done and at what cost? Lancet. 2013;382(9890):452–77. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.