War exposure and changes in eating behaviours in Ukrainian school‐aged children: A cross‐sectional online survey

Abstract

The war in Ukraine has exposed children to extremely high levels of acute and chronic stressors, which can impact their eating behaviour (EB). We aimed to determine the prevalence of war‐induced, stress‐related disruptions in EB of Ukrainian children. We conducted a cross‐sectional online survey among parents of 5‐ to 17‐year‐old children, who had experienced the war in Ukraine in February–May 2023. Guardians reported their child's various war exposure changes in EBs using a modified version of the Child Eating Behaviour Questionnaire. We assessed associations between total and medium‐term EB changes and age, sex, and war exposure using bivariate correlations and χ2 tests. Logistic regression models were fitted to explore the associations between socio‐demographic factors, war exposure and frequency of EB changes. Of the 4854 children, 63% had changes in EBs during the war. The most common EB changes included food cravings (38%), food fussiness (37%), and aversion to certain foods (29%). Of the reported EB changes, 40% were medium‐term, lasting over a month, and related to altered attitudes towards food. Food insecurity (adjusted OR 2.35, 95% CI: 1.76–3.14), and displacement (internally 2.01, 1.19–3.42) emerged as the most influential determinants of medium‐term EB changes. The findings underscore a significant and robust association between war‐related exposures and an increased risk of frequent EB changes. As healthy EBs are learned during childhood and have been shown to track into adulthood, the identified disruptions in EB may have medium‐term consequences for the physical and mental health of Ukrainian children.

This study investigates war‐induced stress‐related disruptions in eating behaviours among 5‐ to 17‐year‐old Ukrainian children who had experienced the war. The findings revealed that two‐thirds of children had changes in parent‐reported eating behaviours (EBs), most frequently observed among 5‐ to 10‐year‐olds. Food insecurity and being displaced emerged as the most influential determinants of EB changes.

Key messages

• Two‐thirds of children had changes in parent‐reported eating behaviours (EBs) during the war, with the highest frequency observed among 5‐ to 10‐year‐olds.

• The most common EB changes observed during wartime included food cravings, food fussiness, aversion to certain foods, and decreased appetite.

• More than one‐third of the reported EB changes were medium‐term, lasting over a month, and related to altered attitudes towards food.

• Food insecurity, displacement, and separation from close relatives emerged as the most influential determinants of EB changes.

• The findings underscore a significant and robust association between various war‐related exposures and an increased risk of frequent EB changes.

1. INTRODUCTION

Since February 2022, the extensive invasion of Ukraine by Russia has led to a significant humanitarian crisis (United Nations News, 2023), with massive civilian casualties and the destruction of infrastructure. At least 10,065 civilians, including over 560 children, have been killed and 18 679 injured since the beginning of the war (United Nations News, 2023). More than 5 million Ukrainians have been internally displaced, and 6.3 million have taken refuge in another country (UNHCR, 2023). Besides the immediate threat posed by war, the mental health consequences of war and migration are enormous. These can include, for example, high prevalence rates of depression or anxiety (Charlson et al., 2019; Hoppen & Morina, 2019; Hoppen et al., 2021; Karatzias et al., 2023; Lytvynenko & König, 2023; Pfeiffer et al., 2023).

For children, exposure to war, conflict zones, migration, and forced displacement can have lifelong impacts on their physical, mental, and social well‐being, as they are especially vulnerable to war‐related stressors (UNICEF, 2021). Childhood adversity can lead to long‐term changes in health‐related behaviour, and physiology, predisposing children to maladaptive behaviours and poor health throughout their lives (Kadir et al., 2019; Samara et al., 2020; Shonkoff et al., 2012). The consequences of war on children are profound and widespread, and a large part of similar studies primarily focus on the psychosocial impacts of armed conflict on children's health (Blackmore et al., 2020; Bürgin et al., 2022; UNHCR, 2023), rather than specific eating behaviours.

Stress reactions can also manifest as unhealthy eating behaviours (EBs), as reported among 8‐ to 18‐year‐olds in a systematic review and meta‐analyses by Hill et al. (2018). Unhealthy EBs described in earlier studies include for example, increased food consumption, compulsive over‐ or under‐eating, changes in taste sensations, and emotional responses to eating (Hill et al., 2018; Hou et al., 2013; Michels et al., 2015). In Chinese adolescents, the effects of emotional stress were examined, finding notable impacts on overeating and food cravings (Hou et al., 2013). The longitudinal study by Michels et al. (2015) examined the relationship between stress and lifestyle over 2 years in 312 Belgian children aged 5–12 years. They found that chronic stress significantly altered children's eating patterns, particularly by stimulating eating in the absence of hunger, which could lead to overweight and obesity (Michels et al., 2015). Unbalanced stress regulation has been found to be more prevalent among children from low socioeconomic status (SES) families, showing how lower SES is associated with higher stress levels and unhealthy eating behaviours in Dutch children (Vliegenthart et al., 2016). However, these studies focus on general stressors in peacetime and do not specifically address children living in conflict zones, where stressors are significantly more acute and chronic. Additionally, they cover a narrow range of eating behaviour dimensions and do not examine their associations with war exposures.

Armed conflicts expose children to exceptionally high levels of acute and chronic stressors, vastly surpassing stress levels in peaceful countries. Earlier studies have reported disrupted EBs during or after armed conflict among Ukrainian adult refugees (Lytvynenko & König, 2023), Ukrainian students (Savelyuk, 2022), Lebanese students (Aoun et al., 2013), and Croatian children (Ajdukovic & Dean, 1998). Disruptions in EB have been described as an increase in the incidence of disordered eating (Aoun et al., 2013; McNulty, 1997), changes in appetite and food intake (Ajdukovic & Dean, 1998; Lytvynenko & König, 2023; Neyland et al., 2020; Savelyuk, 2022), weight loss secondary to starvation, and secondary binge eating disorder (Jacobson et al., 2009). Importantly, the limited research exploring the association between war‐induced stress and EBs primarily focuses on adults and young people. A longitudinal study among 110 displaced 9‐ to 14‐year‐old Croatian children identified a wide range of stress‐related reactions, particularly disordered eating (overly decreased or increased appetite in 32% of the children) (Ajdukovic & DEAN, 1998). In Lebanese students, an overall unhealthy dietary pattern was reported and a change in eating habits during the 2006 July war was associated with an increased risk of eating disorders, with 55% reporting changes post‐conflict (Aoun et al., 2013). Thus, previous research exploring the association between eating behaviours and exposure to war stress has largely focused on adults and young people. To date, no studies have examined the frequency of war‐induced disrupted eating behaviours among school‐aged children. Disordered eating and unhealthy food consumption have been directly linked with greater risk of chronic diseases (Capra et al., 2023; Mescoloto et al., 2023) and obesity (Rousham et al., 2022) and may also contribute to anxiety (Malmir et al., 2023), sleep disturbance (Carvalho et al., 2013), depression, and learning in children (Bürgin et al., 2022; Micali et al., 2015; Michels et al., 2015). That underscores the importance of addressing changes in EB in children.

In the present study, we aimed to evaluate the prevalence of war‐induced changes in EBs among school‐aged Ukrainian children and their potential variations by socio‐demographic factors. In addition, we sought to explore associations between different war exposures and the occurrence of disrupted EB. We assumed that the variability in exposure is driven by the uneven distribution of conflict intensity, displacement patterns, and resource availability. The primary hypothesis of the study was that exposure to war‐induced stress was associated with disruptions in EBs among children, and these associations vary across different age groups, from early childhood to adolescence. It is also hypothesised that socio‐demographic factors, as well as specific war exposures, are associated with children's eating behaviours, frequency, and duration.

2. METHODS

2.1. Study design, participants, and setting

This was a descriptive, cross‐sectional study conducted in Ukraine during the war as an online survey. An a priori power analysis was conducted to determine a sufficient sample size. Based on the earlier literature, we assumed an average prevalence of war‐affected eating disorder rate to be 55% among young people (Aoun et al., 2013). We used a confidence level of 95% and a margin of error of 2%, with 80% power, which indicated the necessary sample size to be 2347 participants. To design the study questionnaire, prior interviews and a prior local survey of parents residing in Kyiv in 2022 were conducted (Gulich & Petrenko, 2022). Interviews with 10 parents of school‐aged children (aged 9–12 years) from two Kyiv schools were carried out by a dietician and a psychologist to determine whether additional constructs were needed to characterise individual differences in eating behaviour and to confirm that the constructs identified from the Children's Eating Behaviour Questionnaire CEBQ (Wardle et al., 2001) were meaningful to parents.

We used a modified version of the CEBQ. The modified constructs were derived from interviews with parents, consultations with a psychology expert, and a preliminary online survey to align with our research goals. Items representing additional constructs (mealtime problems, atypical food preferences and psychosomatic symptoms) were included in the initial questionnaire. In September–October 2022, the questionnaire was piloted with 32 parents of 6‐ to 14‐year‐old children from Kyiv schools (Ukraine), resulting in modifications to some items and wording. Data from the pilot study were used to remove less useful items, discarding those where 90% or more of the parents gave the same response. Based on this, some items were excluded or combined.

Data collection was carried out from 28 February 2023 to 31 May 2023. Invitations and instructions for parents were posted on the official websites of anonymised. Volunteers, including employees of regional laboratories of the State Service of Ukraine for Food Safety and Consumer Protection, shared links in groups of parents of schoolchildren in social networks such as Viber and Telegram. Given the widespread open access invitations to participate in the survey, it is not possible to determine the exact number of those invited, but the online questionnaire was filled in by parents of a total of 5165 children aged 5–17 years living in Ukraine as of 24 February 2022 (the start of the war) with information on family background, war exposure, and potential changes in the child's EB. If a family had multiple children, parents were requested to complete a separate survey for each child. Of the 5165 completed questionnaires, 311 (6%) did not report the age of the child or were reported on behalf of a child who was younger than 5 years or older than 17 years and were thus excluded. The final sample included 4854 children (94%).

2.2. Variables used in the study

The questionnaire contained the following sets of questions used in this study: (1) general socio‐demographic data (age, sex, monthly income per household member), (2) child's location during the war (being in occupied territory, in the immediate vicinity of hostilities, in relatively peaceful territories, internally displaced within Ukraine, displaced abroad), (3) war exposures and distress conditions (e.g., hearing air raid sirens, hearing or seeing explosions, being located in places with long‐term absence of electricity, water, and gas), (4) perceived food insecurity affecting the child's nutrition, and (5) short‐ and long‐term war‐related changes in child's EB and/or attitude towards food (Supporting Information S1: Table S1). Parents indicated whether they had observed any of the listed EB changes in their child by selecting ‘Yes’ or ‘No’. Additionally, they specified the duration of the EB changes (less than 1 month or more). Age groups were condensed to 5–10, 11–13, and 14–17 years. Residence during the war was recategorized as either in relatively peaceful territories or in occupied areas. Displacement status was regrouped into internally displaced, displaced abroad, or stayed at home. The questions were adapted from their original sources, the CEBQ (Wardle et al., 2001), and HFIAS (Coates, 2007) for the research goals and the wartime context. All variables and their modifications are described in Supporting Information S1: Table S1.

2.3. Statistical analysis

Analyses were performed in SPSS version 28.0. We first report descriptive statistics indicating the proportion of children exposed to each war exposure and the proportion of children manifesting EB changes. Next, Chi‐square (χ2) analyses were used to examine the associations between war exposure and EB changes. We assessed whether the total number of EB changes and medium‐term EB changes were associated with age, sex, and war exposure using bivariate correlations and χ2 test. Pearson correlations were computed for all war exposure variables related to EB changes. Finally, univariate and multivariate logistic regression models were used to explore the associations between socio‐demographic factors, war exposure and EB changes. We used two outcomes: having six or more EB changes and two or more medium‐term EB changes, taking the distribution in the number of EB changes into account. Partial correlations, controlling for age, were used to examine relationships between six and more EB changes and two and more medium‐term EB changes and war exposures. We present models with all background and war exposure factors entered separately as well as adjusted models with all background and war exposure factors entered simultaneously.

2.4. Ethics

The current bioethical norms and standards were taken into account in the study design. The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Bioethics Committee of the State Institution ‘O.M. Marzieiev Institute for Public Health’ National Academy of Medical Sciences of Ukraine (protocol no. 4 of 19 July 2022). All participants were informed regarding the objectives and methods of the study. Parents gave their informed consent to participate in the study. All data were collected anonymously under the auspices of (Gulich & Petrenko, 2022).

3. RESULTS

3.1. Socio‐demographic characteristics and war exposure

The socio‐demographic characteristics of the sample are described in Table 1. Most of the children were 5–10 years old (52%) (mean age 10.4 [SD 3.1] years), and half of them (51%) were boys. Most children were from families with average monthly income (70–225€ per household member). The youngest children were more commonly from families with below‐average income levels compared with the older age groups. Overall, 8.6% of the children, more often boys, were from occupied territories and from the immediate vicinity of military actions at the beginning of the war.

Table 1.

Socio‐demographic characteristics and war exposures in a sample of Ukrainian children (n = 4854).

Characteristica	Total sample		Sex			Age
	N	%	Boys n = 2459%	Girls n = 2395%	p value	5–10 years n = 2529%	11–13 years n = 1768%	14–17 years n = 557%	p value
Income level (€/month/household member)
Above average (>225€)a	328	6.9	6.6	7.0	0.89	6.1	7.5	7.9	0.07
Average (70–225€)	2736	57.3	57.5	57.2	0.89	56.9	57.0	60.5	0.07
Below average (<70€)b	1714	35.9	35.9	35.8	0.89	37.0	35.5	31.6	0.07
Residence during the war c
Only in relatively peaceful territories	4117	84.8	84.8	84.8	0.98	85.0	84.5	84.9	0.90
In occupied territories and/or in the Immediate vicinity of hostilities	389	8.0	8.9	7.1	0.03	7.3	8.8	8.6	0.04
Displacement status during the war
Internally displaced (within the borders of Ukraine)	411	8.5	8.5	8.5	0.98	9.1	7.8	7.7	0.26
Displaced abroad	245	5.0	4.9	5.2	0.59	4.6	5.7	5.0	0.32
Perceived food insecurity	244	5.0	5.2	4.8	0.48	4.8	5.2	5.4	0.79
Child's exposure during the war d
Explosions, shelling, air raids, or similar were clearly felt	1400	29.3	29.6	29.1	0.75	28.2	30.9	29.8	0.16
Long‐term absence of electricity, water, gas	1922	40.2	40.1	40.6	0.71	38.0	42.6	43.6	0.003
Forced to leave their own home	599	12.5	12.5	12.7	0.83	13.2	12.2	11.1	0.36
Separated from close relatives for a long time	242	5.1	5.1	4.9	0.70	5.6	4.5	3.9	0.11

^a 225 EUR reflects the average monthly income per household member in Ukraine (7282 UAH/month in 2022) (State Statistics Service of Ukraine, 2022).

^b 70 EUR corresponds to the 2022 subsistence minimum for able‐bodied individuals in Ukraine (2600 UAH/month in 2022) (State Statistics Service of Ukraine, 2022).

^c Missing values: for Residence during the war 7.2% (n = 348), for other variables 0%.

^d Respondent could answer “yes” to multiple options.

Of the total subjects, 8.5% were internally displaced, 5% were displaced abroad, and 12.5% were force‐displaced. About 5% of the children experienced parent‐reported food insecurity, which affected the child's diet for more than a week, but no differences were observed between age and sex groups.

Around 40% of the children, most often in the oldest age group, had experienced long‐term absence of electricity, water, or gas (Table 1). One‐third of children were located in places where explosions, shelling, air raids, or similar were clearly felt. About 5% of the children had been separated from close relatives for a long time.

3.2. Changes in eating behaviours

Disrupted EBs during the war in Ukraine were recorded for 63% of children, and they were more frequent among girls and younger children (Table 2). The most common disrupted EBs were food cravings (38%), food fussiness (37%), aversion to certain foods (29%), and no or decreased appetite (24%). Compared to 11–13‐ and 14–17‐year‐olds, 5–10‐year‐olds showed a higher prevalence of EB changes. Increased appetite (46%) or lack of appetite, food fussiness, aversion to certain foods, and food cravings, as well as the total frequency of EBs, were more common in the younger age group of children (Table 2). The desire to eat at unusual times was, in turn, most common in the oldest age group of children. No or decreased appetite, selective eating, and refusal to eat at regular mealtimes were more typical among girls.

Table 2.

Prevalence of changes in eating behaviour (EB) by age and sex in a self‐selected sample of Ukrainian children (n = 4854).

EB changes	Experienced EB changes % (N)		No changes in EB % (N)	Experienced EB changes,a % (N)
				Boys	Girls	x² (df = 2) p	5–10 years	11–13 years	14–17 years	x² (df = 2) p
	Total EB changesa	Medium‐term EB changes
No or decreased appetite	24.2 (1175)	7.6 (360)	75.8 (3679)	22.7 (558)	25.8 (617)	0.01	26.4 (667)	22.1 (391)	21.0 (117)	<0.01
Fast saturation	18.7 (909)	7.9 (383)	81.3 (3945)	17.9 (440)	19.6 (469)	0.13	20.4 (516)	17.4 (308)	15.3 (85)	<0.01
Delayed saturation	9.3 (453)	3.2 (153)	90.7 (4401)	9.4 (231)	9.3 (222)	0.88	9.1 (231)	9.4 (167)	9.9 (55)	0.85
Selective eating	19.1 (929)	8.7 (424)	80.9 (3925)	17.8 (438)	20.5 (431)	0.02	19.7 (497)	19.8 (350)	14.7 (82)	0.02
Increased appetite	11.9 (579)	4.2 (205)	88.1 (4275)	11.7 (288)	12.2 (291)	0.64	46.3 (268)	41.1 (238)	12.6 (73)	0.01
Food fussiness	36.8 (1784)	20.0 (973)	63.2 (3070)	35.9 (882)	37.7 (902)	0.20	40.2 (1016)	34.1 (603)	29.6 (165)	<0.01
Indiscriminate eating	8.2 (399)	3.7 (180)	91.8 (4455)	8.7 (213)	7.8 (186)	0.26	7.6 (192)	8.5 (151)	10.1 (56)	0.13
Pica	1.6 (76)	0.7 (36)	98.4 (4708)	1.4 (35)	1.7 (41)	0.42	1.7 (43)	1.6 (29)	0.7 (4)	0.23
Aversion to certain foods	28.7 (1394)	15.8 (767)	71.3 (3460)	27.5 (677)	29.91 (717)	0.06	31.9 (808)	25.8 (457)	23.2 (129)	<0.01
Food cravings	38.4 (1864)	21.8 (1057)	61.6 (2990)	37.3 (918)	39.5 (946)	0.12	39.9 (1009)	37.9 (670)	33.2 (185)	0.01
Refusal to eat at usual times	19.6 (953)	7.6 (369)	80.4 (3901)	17.3 (426)	22.0 (527)	<0.001	20.2 (512)	19.1 (338)	18.5 (103)	0.51
Desire to eat at unusual times	15.8 (768)	7.1 (343)	84.2 (4086)	16.1 (396)	15.5 (372)	0.59	13.1 (331)	17.9 (316)	21.7 (121)	<0.01
Desire to chew food constantly increased snacking frequency	15.6 (755)	7.1 (346)	84.4 (4099)	16.0 (394)	15.1 (361)	0.36	14.7 (373)	16.3 (289)	16.7 (93)	0.27
Psychosomatic symptoms	7.2 (350)	2.3 (114)	92.8 (4504)	7.2 (177)	7.2 (173)	0.97	7.4 (186)	7.5 (133)	5.6 (31)	0.27
Total, % (N)	62.9 (3055)	39.7 (1929)	37.1 (1799)	62.3 (1532)	63.6 (1523)	0.35	64.2 (1624)	62.4 (1104)	58.7 (327)	0.04

Abbreviation: N, number of children.

^a Total changes include both short‐ and long‐term EB changes.

Around 40% of children had parent‐reported medium‐term (˃1 month) EB changes, the most common of which were food cravings (21.8%), food fussiness (20.0%), and aversion to certain foods (15.8%). The prevalence of various disrupted EBs, especially those persisting over a period, was more common among younger children (aged 5–10 years). Bivariate correlations explored relationships between age and number of experienced EB changes (mean total EB changes 2.6 (SD: 2.8), mean medium‐term EB changes 1.2 (SD: 1.9). Results showed that age was negatively related to both the total and medium‐term EB changes (r = –0.04, p = 0.01). No significant correlations between sex and EB changes were found.

3.3. Associations between war exposure and changes in eating behaviours

When examined independently, all war exposure variables were significantly associated with six or more EB changes and with two or more medium‐term EB changes (Table 3). In the adjusted models, the highest risk for both all and medium‐term EB changes was observed in the variables' difficulties in providing food and being displaced. All associations between various war exposures and EB changes retained their significance in the adjusted models. In contrast, among socio‐demographic factors, no significant associations were observed in the adjusted models (Table 3). Both adjusted multinomial logistic models were significant (χ²(14) = 144.6 for Model 1 and χ²(14) = 185.8 for Model 2, p < 0.001) and correctly classified 84% of cases with six or more EB changes and 73% of cases with two or more medium‐term EB changes.

Table 3.

Associations between socio‐demographic factors, war exposures, and changes in eating behaviours (EBs).

Characteristic/war exposure, (n)	OR (95% CI) for six or more EB changes (univariate model)	Adjusted model 1a	OR (95% CI) for two or more medium‐term EB changes (univariate model)	Adjusted model 2a
Age, years (N  = 4854)
5–10 ‘n = 2529’	1.24 (0.96–1.60)	1.23 (0.93–1.62)	1.20 (0.97–1.48)	1.17 (0.94–1.46)
11–13 ‘n = 1768’	1.19 (0.91–1.56)	1.14 (0.86–1.52)	1.06 (0.85–1.33)	1.00 (0.79–1.26)
14–17 ‘n = 5577	1	1	1	1
Sex (N  = 4854)
‍‍Boy ‘n = 2459’	0.89 (0.76–1.03)	0.88 (0.75–1.03)	0.97 (0.85–1.10)	0.97 (0.84–1.10)
Girl ‘n = 2395’	1	1	1	1
Income level (€/month/household member) (N  = 4854)
Below average < € 70 ‘n = 1739’	1.53 (1.11‐2.10)*	1.54 (1.10–2.14)	1.19 (0.91–1.54)	1.22 (0.93–1.60)
Average € 70 – € 225 ‘n = 2785’	0.86 (0.63–1.18)	0.94 (0.68–1.30)	0.89 (0.69–1.15)	0.95 (0.73–1.24)
‍Above average > € 225 ‘n = 330’	1	1	1	1
War exposure
Residence during the war b (N  = 4506)
In occupied territories and in the immediate vicinity of hostilities ‘n = 389’	2.01 (1.57–2.54)**	1.35 (0.95–1.94)	2.08 (1.68–2.57)**	1.41 (1.03–1.93)*
Only in relatively peaceful territories ‘n = 4117’	1	1	1	1
Displacement status during the war (N  = 4854)
Internally displaced (within the borders of Ukraine) ‘n = 401’	1.79 (1.41–2.28)**	2.31 (1.24–4.27)*	1.97 (1.60–2.44)**	2.01 (1.19–3.42)*
Displaced abroad ‘n = 245’	1.42 (1.03–1.95)*	1.78 (1.04–3.07)*	1.55 (1.18–2.03)*	1.58 (1.0–2.52)*
Stay at home ‘n = 4208’	1	1	1	1
Perceived food insecurity (N  = 4854)
Difficulties in providing food ‘n = 244’	4.19 (3.21–5.47)**	4.02 (2.98–5.42)**	2.56 (1.97–3.31)**	2.35 (1.76–3.14)**
No difficulties in providing food ‘n = 4610’	1	1	1	1
Child's exposure during the war (yes vs. no) (N  = 4854)
Explosions, shelling, air raids, or similar were clearly felt ‘n = 1425’	1.50 (1.28–1.76)**	1.47 (1.22–1.78)**	1.36 (1.19–1.56)**	1.31 (1.12–1.54)*
Long‐term absence of electricity, water, gas ‘n = 1957’	1.19 (1.02–1.39)*	1.38 (1.16–1.63)**	1.23 (1.08–1.40)*	1.39 (1.21–1.60)**
Forced to leave home with a family ‘n = 611’	1.62 (1.32–2.0)**	1.46 (0.99–2.17)	1.81 (1.52–2.16)**	1.68 (1.20–2.34)*
Separated from close relatives for a long time ‘n = 243’	1.77 (1.31–2.39)**	1.61 (1.12–2.31)*	2.05 (1.57–2.66)**	1.74 (1.27–2.36)**

Note: Results presented as Odds Ratios (95% Confidence Interval).

^a Model includes all exposure variables presented in the table.

^b Missing values: for Residence during the war 7.2% (n = 348).

^* p < 0.05;

^** p < 0.001.

4. DISCUSSION

The findings revealed a notably high prevalence of disrupted EB in Ukrainian school‐aged children during the ongoing war, with almost two‐thirds of children, most frequently 5‐ to 10‐year‐olds, having at least one parent‐reported eating disruption. Changes in appetite and attitudes towards certain foods, including food cravings, fussiness, and aversion, as well as altered mealtimes, were observed as common disruptions in EB among the entire cohort of Ukrainian school‐aged children studied. Various factors related to wartime experiences were associated with both short‐term and medium‐term changes in EB, with the most influential determinants being food insecurity, being displaced, and being separated from close relatives for a long time. This study, to the best of our knowledge, is among the first assessments of the prevalence and frequency of a large set of disrupted EBs in school‐aged children affected by war, as well as their connection to war exposures.

Our findings align with results from a few earlier studies on disrupted EB during the war. In Croatia, during the military conflict in 1992, one‐third of the displaced children developed eating disorders (Ajdukovic & Dean, 1998). Numerous meta‐analyses highlight a significant burden of war stress and mental disorders among conflict‐affected and displaced children (Attanayake et al., 2009; Blackmore et al., 2020; Bürgin et al., 2022; Charlson et al., 2019; Kadir et al., 2019; Samara et al., 2020), which can modify EB (Ajdukovic & Dean, 1998; Higgins Neyland et al., 2020), and may serve as predictors for the development of eating disorders in later life (Aoun et al., 2013; Brewerton et al., 2022; Convertino & Mendoza, 2023). Notably, the prevalence of disrupted EB among children during wartime was markedly higher than the prevalence of stress‐related eating disorders in peacetime (López‐Gil et al., 2023).

Stress response may manifest as a broad array of emotional and behavioural reactions, which can lead to adverse health and contribute to the development of a wide range of later disease outcomes (Bürgin et al., 2022). This issue becomes particularly noteworthy owing to the complex nature of war exposures, including immediate stress responses due to direct war activities (air raids, explosions, shelling, and bombing) and their consequences, for example, traumatisation, multiple losses, abrupt changes in family patterns, living with highly distressed adults, malnutrition, distress from forced separation and displacement, and so forth. (Ajdukovic & Dean, 1998; Bürgin et al., 2022). In addition, the children are also affected by indirect war exposures such as disruptions in food supply, absence of usual foods, and limited access to food. In a recent study among Ukrainian parents, there was evidence of a dose–response relationship between war‐related stressors and meeting the criteria for posttraumatic stress disorders (Lytvynenko & König, 2023).

We investigated the duration of alterations in EB among Ukrainian children exposed to war. Short‐term changes were more prevalent than medium‐term changes, although around 40% of the children had medium‐term changes. The occurrence of short‐term changes in EBs during times of stress may reflect the child's emotional response to traumatic events (Bürgin et al., 2022; Hill et al., 2018; Vliegenthart et al., 2016). The prevalence of several disrupted eating habits was more frequent among younger children (5–10 years), whereas the desire to eat at unusual times and selective eating were more frequent among the oldest group of children (14–17 years). Our study confirms younger children's higher vulnerability to war‐related stressors (CEPR, 2022; Sadeh et al., 2008). Despite significant p‐values, the small effect sizes for some EB changes suggest caution in interpreting practical significance. However, for some EB changes, such as food fussiness (40% among 5–10‐year‐olds vs. 30% among 14–17‐year‐olds), the differences of 10% between age groups can be considered clinically significant.

The long‐lasting disruptions in EB can contribute to the formation of poor eating habits and altered food patterns, which are associated with an increased risk of eating disorders (Aoun et al., 2013; Goldschmidt et al., 2008), and can continue to adulthood and contribute to the development of obesity, and diabetes (Capra et al., 2023; Goldschmidt et al., 2008; Malmir et al., 2023; Mescoloto et al., 2023; Micali et al., 2015; Michels et al., 2015; Rousham et al., 2022; Shonkoff et al., 2012). Identified medium‐lasting EB disruptions in our study were most strongly related to food insecurity, living in occupied territories or in the immediate vicinity of hostilities, and being displaced.

In our study, food insecurity was identified as posing the highest risk for various changes in EB. This finding aligns with prior research that has demonstrated that periods of food restriction can lead to altered EB, such as binge eating and increased food cravings (Bartelmeß & Godemann, 2022; Hazzard et al., 2020; Kosmas et al., 2023). The World Food Programme's briefing to the Council highlighted the significant impact of the Russian invasion on food security, both in Ukraine and globally, with restricted market access and challenges in food production. It also stated that severe food insecurity affects one in five Ukrainian families, intensifying near the hostilities (United Nations News, 2023). Indeed, the disruption of food supply, leading to unstructured eating habits and sometimes even forced dieting or restricted access to food, and lack of familiar foods for children, are stress factors that add to the other detrimental impacts of war. This has led to significant food supply challenges, mandated diet modifications, and alterations in typical food patterns for many individuals, including school‐aged children—a life stage marked by heightened vulnerability to eating disorders (Bürgin et al., 2022; Hill et al., 2018; Micali et al., 2015; Sadeh et al., 2008; Shonkoff et al., 2012). Food insecurity has a substantial impact on an individual's dietary habits and food choices, limiting the diversity of foods (Bartelmeß & Godemann, 2022). Individuals experiencing food insecurity are more likely to overeat or undereat depending on food availability and may consequently be more responsive to the food environment (Kosmas et al., 2023). Furthermore, parents in food‐insecure environments experience elevated psychosocial stress, impacting children who may exhibit varied appetitive behaviours (Eagleton et al., 2022). Food insecurity may contribute to obesity (Eagleton et al., 2022) and health inequalities (Darmon & Drewnowski, 2008) and can lead to social exclusion. Our findings reveal a higher risk for six or more frequent eating disruptions among children from families with below‐average income, consistent with previous reports (Darmon & Drewnowski, 2008; Eagleton et al., 2022).

This study has several limitations that need to be addressed. Firstly, the self‐selected sample is not representative of all school‐aged children in Ukraine, limiting the generalisability of these findings. Given the online data collection method, it is probable that families experiencing severe trauma, particularly in occupied regions or near conflict zones, were probably less likely to engage. Additionally, the potential bias wherein parents who observe more severe behavioural changes are more likely to report stressors must be considered. Unfortunately, due to the ongoing war in the country, records of schools and parent groups invited to complete the questionnaire were not maintained. However, our sample was still diverse, encompassing children from various regions of Ukraine. Although our sample size was adequate for the intended analysis, there was a larger proportion of primary‐school and secondary‐school children; further consideration of the war‐induced EB disruptions in high‐school children would be interesting. Secondly, we assessed EB disruptions in relation to children's experiences during the ongoing war. We did not assess disordered EBs before the war, and thus, we cannot make general statements about changes due to war. Moreover, the reported associations might reflect immediate responses to current conditions rather than long‐term behavioural adaptations. Periods of acute food insecurity observed during the survey might have amplified certain EB changes. As a descriptive study based on self‐reported data, the identified factors are more applicable for hypothesis generation, rather than providing conclusive evidence. In this retrospective study, memory bias could potentially understate the prevalence of certain factors. Finally, while the list of war exposures that we examined was comprehensive, it might not have covered all significant events affecting EBs.

The major drawback of our study is the use of non‐validated instruments to measure food‐related behaviours. While developing our questionnaire, we considered various validated instruments (e.g., Braet & Van Strien, 1997; Bryant et al., 2018; Wardle et al., 2001). Earlier instruments, developed primarily to assess eating behaviours related to obesity, do not encompass the broader range of eating behaviours and stress reactions relevant in a war context. Moreover, consideration of ethical recommendations and suggestions of the Bioethics Committee led us to shorten the questionnaire to minimise stress on participants. It thus remains important to develop and validate a specific version of the war‐affected eating behaviour changes for children and adolescents to better characterise their eating behaviour traits and evaluate the impact of war stress in this population. Despite lacking prior validation, our tool was developed through discussions with parents and experts, along with pilot testing, to ensure consistency and reliability. Although using validated instruments could have offered a comparative foundation, our approach specifically targeted the unique needs and behaviours of the children in our study due to their extraordinary circumstances. We aim to validate our new constructs in future research, incorporating longitudinal analyses to strengthen and broaden the applicability of our findings.

5. CONCLUSION

Despite the limitations, this study holds crucial social implications for Ukraine and other countries facing military conflict, as it sheds light on the behavioural consequences of war. Our findings revealed that two‐thirds of Ukrainian children have changes in parent‐reported EBs, most frequently observed among 5‐ to 10‐year‐olds. More than one‐third of the EB changes were medium‐term, lasting over a month, and associated with altered attitudes towards food. The findings underscore a significant and robust association between various war‐related exposures and increased risk of frequent EB changes. Food insecurity and being displaced emerged as the most influential determinants of EB changes. The medium‐term disruptions were most strongly related to war exposure. These findings contribute to our understanding of the situation and can aid in the development of potential interventions for promoting healthy EB among children and adolescents in extremely stressful environments.

AUTHOR CONTRIBUTIONS

Maria Gulich and Olena Petrenko developed the study protocol and collected the data. Dina Fedorova curated the data, analysed the data, and produced the first draft of the manuscript. Maijaliisa Erkkola and Henna Vepsäläinen supervised the work and reviewed and edited the manuscript. All authors provided critical revisions and approved the final version of the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors report no conflicts of interest.

Supporting information

Supporting information.

Gulich, M. , Fedorova, D. , Petrenko, O. , Vepsäläinen, H. , & Erkkola, M. (2025). War exposure and changes in eating behaviours in Ukrainian school‐aged children: A cross‐sectional online survey. Maternal & Child Nutrition, 21, e13729. 10.1111/mcn.13729

DATA AVAILABILITY STATEMENT

Data that support the findings of this study are available from the corresponding author upon reasonable request.