Food security and its impact on growth among Sri Lankan children under five during the economic crisis in 2022

Abstract

The recent economic recession has reportedly worsened food insecurity in Sri Lanka. We assessed food insecurity and its impact on the growth of children aged 6–59 months through a community-based, descriptive, cross-sectional study conducted in 2022. Food insecurity was measured using the Household Food Insecurity Access Scale, and anthropometric measurements (weight, length/height) were taken using standard techniques. Weight faltering was defined as inadequate or no weight gain (flattening) or a drop in weight gain, based on the trajectory of the weight curve over two points six months apart. Additionally, food insecurity, stunting, and wasting during crisis were compared with pre-crisis data. The study included 832 children, with half of the households experiencing moderate to severe food insecurity. One-fifth of the children showed growth faltering during the crisis, and 8.9% of those with growth faltering crossed one centile line (0.67 SD). Among children facing food insecurity, the rate of growth faltering had doubled, with the residential sector (i.e., estate) tripling the risk of weight faltering. The rates of food insecurity and growth issues were notably higher during the crisis compared to pre-crisis data. This study highlights the worsening food insecurity and its significant impact on growth faltering in children under five during the 2022 economic recession.

Background

In 2019, Sri Lanka was 66th of 113 countries in the Global Food Security Index [1]. However, following the unforeseen COVID-19 pandemic, Sri Lanka was hit by its worst economic recession, affecting all layers of society [2]. Not only did the COVID-19 pandemic claim many lives, but it also severely disrupted agricultural activities, the country’s primary source of food supply [3]. Prices of essential food items reached a record high in September 2022, pushing the year-on-year food inflation rate to nearly 95% [2].

Although Sri Lanka has achieved better maternal and child mortality statistics than other developing countries in the Asian subcontinent, undernutrition remains a significant public health concern [4, 5]. Many factors cause undernutrition in children, and one main reason is food insecurity [6]. The UNICEF conceptual framework describes poor dietary intake as a leading cause of undernutrition [6]. Children and pregnant women are more vulnerable to undernutrition in the presence of food insecurity than other segments of the population. As per the last demographic health survey in 2016, rates of wasting and stunting were recorded as 15% and 17%, respectively [7]. Another analysis of data obtained through island-wide routine growth monitoring assessment carried out by the Family Health Bureau of Sri Lanka reported a lower but increasing trend of growth failure indicators from 2021 (wasting-8.2%, and stunting-7.4%) to 2023 (wasting-10% and stunting-9.2%) which was attributed to the COVID-19 pandemic and the unprecedented economic recession [8].

The recent economic recession has worsened food security in Sri Lanka. In October 2022, 54% of the households experienced acute food insecurity [9]. Due to rising food prices and reduced purchasing power, more households ate less preferred, cheaper food and limited portion sizes, which had the worst effect on low-income people.

Before the recession, limited regional studies reported higher wasting, stunting, and underweight rates among children from food-insecure households [10]. As mentioned above, many surveys have reported deterioration of food insecurity during the economic recession. However, the effect of the deterioration of food insecurity on children’s growth has not been evaluated. Thus, we aimed to assess food security and its impact on growth in children aged 6–59 months during the economic recession in 2022. Understanding these dynamics is essential for implementing targeted interventions to lessen the effects of the economic recession on child nutrition.

Methods

A community-based descriptive cross-sectional study was conducted among children aged 6–59 months from September to November 2022. Participants were selected using multi-stage stratified cluster sampling from 16 Medical Officer of Health (MOH) areas in four selected districts (Gampaha, Ratnapura, Jaffna, and Nuwara Eliya) in Sri Lanka. These districts were purposefully selected to ensure adequate representation of the residential sectors (estate, rural and urban sectors) of the population. Based on the demographic and economic characteristics, MOH areas belonged to these districts were divided into three sectors: estate, rural and urban, to ensure adequate representation of the residential sectors of the population. Plantation areas spanning over 20 acres and employing at least ten residential labourers are classified as the estate sector. Other residential areas not belonging to the urban/estate sector are classified as rural.

Exclusion Criteria were children with a severe physical disability where standard growth measurements cannot be done accurately, children who have received in-patient care during the past month, children who have experienced a diarrheal illness or other febrile illness lasting more than five days within the last two weeks, irregular clinic visits and not having weight measurement during the defined period and having a complex medical co-morbidity or disability.

Written informed consent was taken from the caregiver prior to the study. Voluntary participation was assured, and they were informed they could withdraw at any stage.

Sample size calculation

The sample size was calculated using two methods. As per the method based on Lwanga & Lemeshow 1991, with an expected prevalence of under 5 wasting as 8.6% [8], the margin of error of 0.05, n = 121 for each selected district. When design effect (1.5) and non-response rate (10%) were considered, the total sample size reached 806. The second method used a formula to compare two proportions to calculate the association between malnutrition and food insecurity [11, 12]. A recent study in Bangladesh evaluated the impact of the COVID-19 lockdown on the nutrition and food security of low-income households using the Household Food Insecurity Access Scale (HFIAS). Among children with wasting, 23.4% had no or mild food insecurity, and 32.4% had moderate to severe food insecurity [13]. Accordingly, when the formula was applied, the sample size was 822 when α was 0.05, and the power of the study was 80%. As both sample sizes were similar, using four districts with 208 subjects (4 clusters of 52) in each resulted in a total of 832 subjects. Sample size calculation was performed using PASS statistical software.

Sampling technique

Children were recruited using a stratified cluster sampling method. Four districts were selected based on their predominant sectors: predominantly rural (Jaffna and Ratnapura), predominantly urban (Gampaha), and predominantly estate (Nuwara Eliya). In the next stage, four Medical Officer of Health (MOH) areas were listed for each district. These MOH areas were then classified into three sectors, and four MOH areas were randomly selected from each district. The MOH area is the smallest administrative unit in public health in Sri Lanka and one MOH area was considered as a cluster. Sector classification information was obtained from the Department of Census & Statistics. In each cluster, investigators enrolled every third participant who visited the clinic on that day. This process was repeated three days a week per cluster, with the selected days determined by the clinic schedule of each cluster. The enrollment continued until the sample size was achieved.

Study instruments

An interviewer-administered questionnaire was used for data collection (socio-demographic details, anthropometry, egg/flesh food consumption during the previous 24 hours, and household food insecurity). Household Food Insecurity Access Scale (HFIAS) is a valid tool to use in a developing country context and measures food insecurity with cross-cultural equivalency [14]. It is a continuous measure of assessing past 4-week food insecurity. The HFIAS questionnaire focuses on three domains of food insecurity: Concern and uncertainty about the family’s food supply, change in diet quality, and amount of food consumed. These responses were recorded based on the frequency of occurrence - ‘no occurrence’, ‘rarely’, ‘sometimes’ and ‘often’. These responses were scored according to the manual. The score ranged from 0 to 27 (The higher the score, lower the family household food insecurity). Thus, food secure (HFIAS 0–1), mildly food insecure (HFIAS 2–7), moderately food insecure (HFIAS 8–11), and severely food insecure (HFIAS > 11) households were identified. HFAIS has been previously translated and validated in both Sinhalese and Tamil languages [15–17].

Electronic seca 874 (seca GmbH & Co. KG.) scale was used for weight measurement with 50 g precision. Weight was measured with minimum clothes, and the scales were calibrated each day before use. Length was measured to last completed 1 mm using a portable measuring board (S0114530 Length/height board, baby/child, SET-2). The data were collected by medical graduates awaiting their internship placements. All data collectors were trained in administering the questionnaire and anthropometric measurements.

Definitions

Presence of weight faltering was defined as inadequate or no weight gain (flattening) or a drop in weight gain, based on the trajectory of the weight curve over two time points during the economic recession (a six-month period before the point of recruitment) [18]. Current weight for age, height/length for age, and weight for height were interpreted and underweight, stunting and wasting were defined per the World Health Organization (WHO) recommendations [19]. Median household income was taken as 53,333 LKR [20].

Statistical analysis

Statistical software SPSS 22.0 was used for data analysis. Simple and multiple binary logistic regressions were used to assess the determinants of weight faltering, stunting and wasting. For regression analysis, the dependent variables were weight faltering, stunting and wasting. The measurements taken at the time of recruitment were considered in assessing stunting and wasting. The observation of growth curve in the previous six months as described in the definition was taken as growth faltering. The residential sector was collapsed into a dichotomous variable (Estate vs. non-estate) since there was no difference between urban and rural sectors. The multicollinearity of independent variables was assessed with pairwise correlation coefficients between predictors and variance inflation factor (VIF). The father’s educational level and father’s employment status were not included in the model since there was multicollinearity with other socioeconomic indicators. All variables with a p value < 0.2 in the univariate analysis and variables with a-priori knowledge were considered in multivariate regression analysis (e.g., income, parent educational level). Goodness-of-Fit for Logistic Regression was assessed using the Hosmer-Lemeshow test. Results were given as crude and adjusted odds ratios with 95% confidence intervals. P-values < 0.05 were considered statistically significant.

The study compared the region-wise data from previous surveys on food security and growth problems in children conducted before the crisis in Sri Lanka with the findings of the current study [20–22]. Since province-wise information on food security was available, the study extrapolated district-wise data based on the districts belonging to each province.

Results

A total of 832 children were included in the analysis, with a 95.9% response rate. Thirty-four were excluded due to infrequent clinic visits without weight measurements in the previous year (n = 26), incomplete interviews (n = 5), and implausible growth measurements (n = 3). Two-thirds of the children were less than 24 months old (Table 1). A significant proportion had low birth weight. The majority of parents passed their ordinary level exam. Household income was less than the median in the majority.

Table 1.

Baseline characteristics

Variable	Categories	Frequency (%)
Age (months)	< 24 months	549 (66)
	≥ 24 months	283 (34)
Gender	Male	418 (50.2)
	Female	414 (49.8)
Mother’s Education Level	Below O/L	204 (24.5)
	O/L and above	628 (75.5)
Father’s Education Level	Below O/L	202 (24.3)
	O/L and above	630 (75.7)
Total household income (LKR)	< Median	569 (68.4)
	≥ Median	263 (31.6)
Father’s employment	Non-skilled	190 (22.8)
	Skilled	642 (77.2)
Sector	Urban	391 (47)
	Rural	294 (35.3)
	Estate	147 (17.7)
Birth weight (kg)	< 2.500 kg	146 (17.5)
	≥ 2.500 kg	686 (82.5)
Diarrhoeal episodes (previous 6 m)	Yes	36 (4.3)
HFIAS, Overall	No/ Mild food insecurity	440 (52.9)
	Moderate food insecurity	238 (28.6)
	Severe food insecurity	154 (18.5)
Growth problems	Weight faltering	163 (19.6)
	Wasting	79 (9.5)
	Stunting	110 (13.2)
	Underweight	137 (16.5)
	Normal	547 (65.7)
Weight faltering	Nuwara Eliya	66 (31.7)
	Gampaha	41 (19.7)
	Jaffna	25 (12)
	Ratnapura	31 (14.9)

Food insecurity and growth problems

47% (n = 392) of households had moderate to severe food insecurity, and one-third of them (n = 154, 39.3%) had severe food insecurity. A statistically significant difference in food insecurity was found between residential districts (p < 0.001) but not between residential sectors (p = 0.9). One-fifth had growth faltering, while 9.5% were wasted, 13.2% were stunted and 16.5% had underweight. Among the children with growth faltering, 8.9% crossed one centile line (0.67 SD).

Based on two surveys conducted in 2021, a comparison of food insecurity and growth problems are given in Tables 2 and 3 [20–22]. In these surveys, food security was measured by the Global Food Insecurity Experience Scale. Growth parameters in children in the Ratnapura district during the annual Nutrition Month Growth Assessment 2021 have not been reported [22]. A higher rate of food insecurity was noted in all four studied districts compared to pre-crisis data, with a notable difference between the Gampaha and Ratnapura districts. Regarding growth parameters, except for the stunting rate in the Jaffna district, all other parameters in other districts are higher than in the pre-crisis period. In the present study, the stunting rate in Nuwara Eliya is almost twice the reported pre-crisis value. Weight faltering was highest in Nuwara Eliya district (31.7%) compared to Gampaha (19.7%), Ratnapura (14.9%) and Jaffna (12%).

Table 2.

Food insecurity: comparison of the current study findings and previous pre-crisis surveys

Residential district	Moderate/Severe insecurity (%)		p value	Severe insecurity (%)		p value
	Current study	*Pre-recession (2021)		Current study	*Pre-recession (2021)
Gampaha	60.1	13.6	< 0.001	16.3	0.7	< 0.001
Jaffna	27.9	25.6	0.4	13.9	2.5	< 0.001
Nuwara Eliya	48.1	38.3	0.008	19.7	5.5	< 0.001
Ratnapura	52.4	17.3	< 0.001	24.0	2.3	< 0.001

*Based on surveys conducted in 2021 [22, 23]

Table 3.

Growth problems: comparison of the current study findings and previous pre-crisis surveys

Residential district	Stunting (%)			Wasting (%)			Underweight (%)
	Current study	*Pre-recession (2021)	p value	Current study	*Pre-recession (2021)	p value	Current study	*Pre-recession (2021)	p value
Gampaha	8.7	05	0.03	8.2	7.3	0.6	12.1	9.6	0.01
Jaffna	6.7	7.1	0.8	8.2	7.8	0.8	9.6	11	0.19
Nuwara Eliya	31.3	17.4	< 0.001	10.5	9.7	0.7	27.6	20	< 0.001
Ratnapura	5.8	NR	-	11.1	NR	-	16.3	-	-

*Based on surveys conducted in 2021 [22, 23]. NR: Not recorded

Determinants of weight faltering during the economic recession

Simple logistic regression indicated that food insecurity, and residential sector were significantly associated with weight faltering. Residents from estate sector, had more than three times the risk of weight faltering (OR: 3.41, 95%CI: 2.06, 5.610, p-value:<0.001). Due to multicollinearity, only the sector variable was included in the final model, and the associations remained the same after adjustment to covariates (Table 4). The Hosmer-Lemeshow test showed that the model fitted the data well, p = 0.681.

Table 4.

Simple and multivariable logistic regression for determinants of weight faltering during the economic recession

	Crude OR	95% C.I.		P value	Adjusted OR	95% C.I.		P value
Food insecurity
No/mild insecurity	Ref				Ref
Moderate insecurity	1.32	0.882	1.962	0.2	1.45	0.945	2.229	0.09
Severe insecurity	1.68	1.079	2.606	0.02	1.85	1.133	3.013	0.01
Age
6–24 m	Ref				Ref
≥24–59 m	1.13	0.780	1.625	0.5	1.08	0.738	1.585	0.7
Gender (Male)	0.86	0.607	1.206	0.4	0.85	0.589	1.219	0.3
Birthweight
<2.5 kg	Ref				Ref
≥2.5 kg	1.25	0.811	1.924	0.3	1.13	0.710	1.782	0.6
Income
< Median	Ref				Ref
≥ Median	1.18	0.808	1.713	0.4	0.85	0.559	1.299	0.5
Mother’s education
Below O/L	Ref				Ref
O/L and above	0.76	0.516	1.110	0.2	1.18	0.749	1.873	0.5
Diarrhoeal illness (Yes)	2.14	1.045	4.367	0.03	2.14	1.000	4.561	0.05
Residential sector
Urban sector	Ref				Ref
Rural sector	0.93	0.616	1.405	0.7	0.90	0.592	1.381	0.6
Estate sector	2.75	1.785	4.222	< 0.001	3.41	2.069	5.610	< 0.001

aOR: Adjusted Odds Ratio, cOR: Crude Odds Ratio

Determinants of stunting

Simple logistic regression indicated that severe food insecurity, children from the estate sector, and low birth weight were likely to be stunted. Also, higher household income and maternal education were protective against stunting. Children aged 24–59 months had a higher risk of stunting than 6–24 months. Only the child’s age, gender, birthweight, district, moderate/severe food insecurity and household income remained significant in the multivariable logistic regression (p < 0.05). Children from Nuwara Eliya district had a fourfold higher risk of being stunted. (Table 5). The Hosmer-Lemeshow test showed that the model fitted the data well, p = 0.550.

Table 5.

Simple and multivariable logistic regression for determinants of stunting

	Crude OR	95% C.I.		P value	Adjusted OR	95% C.I.		P value
Food insecurity
No/mild insecurity	Ref				Ref
Moderate insecurity	1.12	0.678	1.833	0.7	0.92	0.521	1.638	0.8
Severe insecurity	2.37	1.457	3.853	0.001	1.87	1.056	3.313	0.03
Age
6–24 m	Ref				Ref
≥24–59 m	2.03	1.333	3.079	0.001	2.01	1.264	3.198	0.003
Gender (Male)	1.58	1.051	2.384	0.03	1.70	1.066	2.715	0.03
Birthweight
<2.5 kg	Ref				Ref
≥2.5 kg	3.05	1.955	4.751	< 0.001	2.98	1.813	4.906	< 0.001
Income
< Median	Ref				Ref
≥ Median	0.29	0.161	0.515	< 0.001	0.45	0.233	0.865	0.02
Mother’s education
Below O/L	Ref				Ref
O/L and above	0.32	0.211	0.484	< 0.001	0.66	0.398	1.101	0.1
Diarrhoeal illness (Yes)	1.35	0.547	3.312	0.5	2.14	0.799	5.729	0.1
Residential sector
Urban sector	Ref				Ref
Rural sector	0.98	0.569	1.675	0.9	0.97	0.552	1.715	0.9
Estate sector	5.41	3.316	8.835	< 0.001	4.03	2.263	7.166	< 0.001

aOR: Adjusted Odds Ratio, cOR: Crude Odds Ratio

Determinants of wasting

Simple logistic regression indicated that children with moderate and severe food insecurity, increasing age and low birth weight had a higher risk of wasting. These associations remained the same after the adjustment to all covariates (Table 6). The Hosmer-Lemeshow test showed that the model fitted the data well, p = 0.480.

Table 6.

Simple and multivariable logistic regression for determinants of wasting

	CrudeOR	95% C.I.		P value	Adjusted OR	95% C.I.		P value
Food insecurity
No/mild insecurity	Ref				Ref
Moderate insecurity	2.21	1.269	3.848	0.005	2.26	1.229	4.141	0.009
Severe insecurity	2.94	1.632	5.297	0.003	2.89	1.489	5.638	0.002
Age
6–24 m	Ref				Ref
≥24–59 m	1.59	0.975	2.586	< 0.001	1.69	1.008	2.840	0.04
Gender (Male)	0.83	0.522	1.323	0.4	0.66	0.397	1.110	0.1
Birthweight
<2.5 kg	Ref				Ref
≥2.5 kg	4.06	2.490	6.631	< 0.001	4.33	2.554	7.341	< 0.001
Income
< Median	Ref				Ref
≥ Median	1.22	0.729	2.042	0.4	0.92	0.509	1.661	0.7
Mother’s education
Below O/L	Ref				Ref
O/L and above	0.67	0.407	1.114	0.1	0.76	0.411	1.415	0.3
Diarrhoeal illness (Yes)	1.20	0.414	3.490	0.7	1.15	0.359	3.707	0.9
Residential sector
Urban sector	Ref				Ref
Rural sector	0.88	0.531	1.483	0.6	0.87	0.502	1.517	0.6
Estate sector	0.78	0.397	1.532	0.5	0.55	0.245	1.247	0.2

aOR: Adjusted Odds Ratio, cOR: Crude Odds Ratio

Discussion

This study evaluated child growth issues within the context of the economic recession, which was one key area to explore to assess the impact. In addition, we compared our findings with the existing pre-cr251-2isis data. Among children from families experiencing food insecurity, the rate of growth faltering has doubled. Multiple household and contextual factors were associated with growth failure, which may be directly or indirectly associated with the economic recession.

In Sri Lanka, the prevalence of moderate to severe food insecurity had been increased compared to pre-crisis data in 2021 [21]. The reported food insecurity level in the urbanized district (Gampaha) was striking higher than pre-crisis data (60% vs. 13.6%) [21]. The likely reasons could be the following. Urban households generally allocate a more significant portion of their spending to non-food items rather than food [23]. During the economic crisis, expenditure on non-food items may have increased markedly. As shown by surveys, household expenditure on fuel, health, utilities, and education, in addition to food items, increased exponentially. This situation may have significantly impacted food security in the urban sector more than in other sectors. Rural and estate sectors spend on low-priced energy-yielding staples such as rice and yams to meet the family’s minimum energy requirements. At the same time, urban households, with their long working hours, often prefer spending on prepared food rather than cooking [23]. Also, rural households engage in home gardening for alternative food sources. Additionally, compared to other sectors, sharing food items among households in rural areas is common in many communities, ensuring everyone has access to some amount of food, especially during challenging times.

In 2022, during the crisis, comparable to our study findings, another survey using the Household Food Insecurity Access Scale (HFIAS) indicated a doubling of these figures to 54% [9]. Further, as shown in Table 2, in district-wise comparison, food insecurity in each district during a crisis is higher than in pre-crisis data. Although these surveys use different tools, food insecurity has likely worsened in all districts due to the economic downturn and inflation.

However, the significantly higher food insecurity rates in a predominantly rural district (Ratnapura) compared to pre-crisis figures are debatable. For two reasons, pre-crisis food insecurity may have risen more in Ratnapura than in Jaffna (rural). Firstly, the loss of agricultural land due to increased gem mining activities in the Ratnapura area might have reduced the available space for crop cultivation, further threatening food security and the livelihoods of local farmers [24]. This situation may have been exaggerated by the government’s sudden ban on chemical fertilizer imports during this period, with a more than 50% reduction in crop yield [25].

A higher rise in food insecurity was seen in Nuwara Eliya compared to other districts, likely due to high rate of poverty [23]. However, a marginal rise was noted in our study and this could be due to presence of emergency food supplementation programs operated during crisis as safety nets by government and non-government organizations in Nuwara Eliya. These safety nets were not considered in our evaluation. These programs may have mitigated the effects of food insecurity to some degree, obscuring the true extent of food insecurity in this region.

When comparing pre-crisis data with the study findings on stunting, a significant increase in stunting was observed in all districts. However, the rise was not significant for wasting (Table 3). The most likely explanation is that increased food insecurity led to replacing expensive, protein-rich foods with a carbohydrate-based diet, reducing the overall quality of meals. Previous studies have shown that a low-protein diet is associated with stunting [27, 28].

In our study, weight faltering, stunting and wasting showed a positive association with food insecurity. Weight faltering was associated only with severe food insecurity, possibly due to adults consuming fewer quantities and prioritizing children’s diets, which is generally the predictable Asian culture. In contrast, in severe food insecurity, there is a significant food shortage to the extent that individuals, including children, cannot access or be provided with adequate food for their nutritional needs. While wasting was associated with both moderate and severe food insecurity, stunting was linked only with severe food insecurity. It can be assumed that food insecurity and other factors, such as genetic, maternal/prenatal, and environmental, influence linear growth in children [29].

We have reported factors other than food insecurity associated with growth problems. The estate population showed a higher rate of growth faltering and stunting. Other than food security, it is very likely that multiple factors such as low literacy level may have contributed to growth problems in the estate sector [30]. Although stunting was a significant issue among the estate population, wasting was not significantly associated. One likely explanation is that because children were severely stunted, detecting wasting was challenging when there was a less pronounced deficit in weight for height compared to the deficit in height. Therefore, in populations where stunting is prevalent, wasting may be underestimated or not detected as readily, giving the false impression that wasting is less prevalent.

It is also important to note that birthweight influences growth problems, both stunting and wasting. Several studies in Sri Lanka have shown a link between birth weight and child undernutrition, indicating that low birth weight is the most critical factor contributing to stunting, wasting, and underweight in children under five years old [31]. Thus, interventions should be prioritized and strengthened, focusing on antenatal and postnatal care in the vulnerable segments.

In addition, increasing age, gender and low-income were associated with stunting. The association between increasing age and stunting has been shown previously [22]. This pattern suggests that as children grow older, the factors contributing to undernutrition become more pronounced. The cumulative effects of inadequate nutrition and socioeconomic challenges may intensify with age, thereby increasing the risk of stunting and other forms of undernutrition. Males were at risk of stunting than females. Our findings for all children under the age of five align with previous research, which typically shows a higher prevalence of stunting among boys compared to girls [32, 33]. The comparative disadvantage of young boys aligns with existing evidence; it may be partly due to immunological factors impacting the severity of infectious diseases and sex-specific hormones that increase male susceptibility to undernutrition under 30 months [34, 35].

Understanding the relationship between food insecurity and other factors with growth failure is vital for updating policy and program development related to child growth and well-being during economic instability. The specific determinants of growth failure could be utilized to implement targeted interventions to mitigate the impact of economic crises on child health. However, there is a need for more information and research on the impacts of the crisis on child nutrition. Intervention via safety net programmes is important. Vulnerable households need support on coping strategies in addition to their efforts.

The study findings should be interpreted with the following limitations. The study sample may partially represent the whole child population of Sri Lanka since the sample was drawn from the state health sector clinics. It may have yet to adequately cover the upper socioeconomic strata as they are followed up at private-sector clinics. When food insecurity was assessed with the HFIAS, caregivers may have given answers that they thought were socially acceptable or desirable rather than reflecting their actual practices. This bias could affect the accuracy of the data. We relied on the weight measurements collected by field staff in community clinics before the onset of the economic recession, which could have introduced inaccuracies. The wealth index could have been a better economic indicator than household income since it would consider assets and amenities in addition to income, and could have provided a more comprehensive measure of economic status.

Conclusion

Our study found that food insecurity has worsened and has impacted growth faltering in children under five during an economic recession. It also highlights the complex interplay of factors contributing to growth failure, in addition to food security, such as gender, low birth weight, low income, and living in the estate sector. Thus, emphasizing the need for targeted interventions and policy measures focusing on the vulnerable population segments to improve food security in the country.

Author contributions

D.V., H.S. and G.L. conceptualized the study.D.V., H.S., J.R., G.S., G.U.S., A.D., P.C., and G.L., involved in designing the study.D.V. involved in funding acquisition.G.L. wrote the original draft.D.V., H.S., J.R., G.S., G.U.S., A.D., P.C., M.S, and G.L., reviewed the manuscript.

Funding

This research was funded by the Sri Lanka College of Paediatricians.

Data availability

Data will be available from the corresponding author upon a reasonable request.

Declarations

Ethical approval

The Ethics Review Committee of Sri Lanka College of Paediatricians approved the protocol (SLCP/ERC/2022/26) on 26.07.2022. The caregivers gave written consent for review and signature before starting interviews and the study was conducted in accordance with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.