Comparison of eating behaviour by relative weight change status of young adults throughout the COVID-19 pandemic

Abstract

Background and purpose

The enforcement of nationwide lockdowns has worsened the obesity epidemic in Malaysia. Therefore, this study aims to compare eating behaviour by relative weight change status among young adults in Malaysia throughout the COVID-19 pandemic.

Design/methodology/approach

Socio-demographics, body height, pre-pandemic body weight, and post-lockdown body weight were self-reported by the young adults. The eating behaviour was assessed using the validated Three-Factor Eating Questionnaire-Revised 18-item (TFEQ-R18). The mean difference in eating behaviour subscales was analysed using MANCOVA with a Bonferroni-adjusted post-hoc test.

Findings

One-third of the young adults put on body weight due to the lockdowns, with an average relative weight gain of 12.44 ± 9.67%. Conversely, about one-fifth of the young adults reported having lighter body weight in the post-lockdown pandemic phase, with an average relative weight loss of 10.10 ± 4.66%. Young adults in the weight loss category had a significantly higher standardised raw score in cognitive restraints than those in the sustained weight and weight gain categories. In addition, young adults in the weight gain category had statistically higher standardised raw scores in uncontrolled eating and emotional eating compared to those in the sustained weight category. Weight trajectory during the COVID-19 pandemic is linked to disordered eating behaviour among young adults in Malaysia.

Originality/value

The findings presented in this study can be potentially valuable in formulating weight management strategies in the post-COVID-19 era.

1. Introduction

The coronavirus disease 2019 (COVID-19) is a highly contagious disease caused by the SARS-CoV-2 virus [1,2]. In Malaysia, the number of COVID-19 confirmed cases continued to rise at an alarming rate, even though three nationwide lockdowns (also known as the Movement Control Order, MCO) had been implemented. As of April 2022, Malaysia recorded 4.2 million confirmed cases and 35 thousand deaths from COVID-19 [1,2]. From 2020 to 2021, the Federal government of Malaysia took several precautionary measures to mitigate COVID-19 transmission in the nation. One of those includes restricting the movement of people through the enforcement of total lockdown during this unprecedented pandemic.

On June 15, 2021, the National Recovery Plan (NRP) was unveiled to the public to replace the existing MCO [3]. The shift from MCO to NRP was executed due to the rapid COVID-19 vaccine rollout starting in June 2021 [4]. The NRP comprises four phases of pandemic exit strategies to ensure all sectors return to their pre-pandemic levels. The transition from one phase to another relies on three headline indicators: the vaccination rate among the adult population, daily new cases, and utilisation of the Intensive Care Unit (ICU) [5]. Phase 1 of NRP implies a total lockdown with the tightest Standard Operating Procedures (SOPs). In contrast, phase 4 represents a full reopening of all economic sectors and more lenient SOPs [6].

Although all states in Malaysia entered phase 4 of NRP on January 3, 2022, the devastating impacts of COVID-19 on body weight continue to deepen during the post-lockdown pandemic phase. A recent study by Tan et al. [7] pointed out that the prevalence of overweight/obesity in young adults in Malaysia rose from 37.7% during the MCO 3.0 to 39.2% after the lockdown was lifted. The increase in obesity rates was primarily attributed to increased sedentary behaviour, increased refined carbohydrates and fat consumption, a substantial decrease in moderate-to-vigorous physical activity, and poor self-regulation of eating behaviour during the pandemic [[8], [9], [10], [11]]. Therefore, this study aims to compare eating behaviour by relative weight change status among young adults in Malaysia throughout the COVID-19 pandemic.

2. Methodology

2.1. Study design and study population

Data collection was conducted from January 10–31, 2022 (22 months after the first nationwide lockdown). A combination of convenience and snowball sampling approaches was applied to recruit young adults into this study. The questionnaire was hosted on google forms and circulated to the young adults via social media platforms, including Facebook, Twitter, Whatsapp, Instagram, and TikTok. Young adults who responded to the survey invitation were encouraged to share the survey link with someone they knew.

Individuals aged 18–30 years, free from obesity-related comorbidities, accessible to the internet during the data collection period, free from clinically diagnosed eating disorders, and not participating in weight management interventions throughout the pandemic were included in this study. This study was conducted according to the guidelines laid down in the Declaration of Helsinki, and all procedures were approved by the Research Ethics Committee of Management and Science University with the reference number MSU-RMC-02/FR01/12/L1/124. Written informed consent was obtained from all subjects prior to answering the first survey questions. The sample size was determined using the G*Power software version 3.1 (Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany) with a 95% power at α = 0.05 to detect a medium effect size of 0.25 [12]. Hence, a minimum sample size of 54 young adults is required for this study.

2.2. Socio-demographic characteristics of the young adults

Socio-demographics, including gender, age, ethnicity, and marital status, were self-reported by the young adults. In addition, they were also required to disclose their monthly earned income during the data collection period.

2.3. Eating behaviour throughout the COVID-19 pandemic

The eating behaviour of the young adults was assessed using the validated Three-Factor Eating Questionnaire-Revised 18-item (TFEQ-R18) [13]. It comprises 18-item which can be further subcategorised into three eating behaviour subscales: cognitive restraint (6-item), uncontrolled eating (9-item), and emotional eating (3-item). All items in the TFEQ-R18 were coded on a 4-point scale (1–4). The raw scores were obtained by summating item scores of each subscale and were converted to standardised raw scores with a 0–100 scale by using Formula 1:

Formula 1:

$$\text{Standardised}\mathrm{raw}\text{score}=\frac{\mathrm{Raw}\text{score}-\text{Lowest possible}\mathrm{raw}\text{score}}{\text{Possible}\mathrm{raw}\text{score range}}\times 100$$

Higher scores in the respective subscales imply greater cognitive eating, uncontrolled eating, or emotional eating. The internal consistency of TFEQ-R18 was good, with a reported Cronbach's alpha of 0.783.

2.4. Weight change status throughout the COVID-19 pandemic

The current study required young adults to self-report their body height (cm), pre-pandemic body weight, and post-lockdown body weight. The pre-pandemic body weight (kg) was obtained by requesting the respondents to recall their body weight retrospectively before the declaration of MCO 1.0 in Malaysia (body weight during February 2020). With regards to post-lockdown body weight, the respondents were encouraged to measure their current body weight (kg) with a bathroom scale in their capacity. The relative weight change (%) was calculated based on the formula suggested by Tan et al. [14]. It was stratified into three groups: sustained weight (± 4.99% of the pre-pandemic body weight), weight loss (≤ −5% of the pre-pandemic body weight) and weight gain (≥ 5% of the pre-pandemic body weight). The pre-pandemic and post-lockdown BMI were also quantified and classified into underweight (< 18.5 kg/m²), normal (18.5–22.9 kg/m²), overweight (23.0–24.9 kg/m²), and obese (> 25.0 kg/m²) based on the Asia-Pacific cut-off points [15].

2.5. Statistical analysis

Data analysis was conducted using IBM SPSS version 26.0 (IBM Corp, Armonk, NY). Categorical variables were expressed in frequency and percentage, while continuous variables were reported in mean and standard deviation. The mean difference between pre-pandemic BMI and post-lockdown BMI was analysed using paired samples t-test. On the other hand, the relationship between eating behaviour subscales and potential confounding variables (gender, age, ethnicity, marital status, and monthly earned income) was assessed using Pearson's correlation test. Variables that portray statistically significant correlation in Pearson's correlation test were subsequently included as covariates in the Multivariate Analysis of Covariance (MANCOVA) analysis. The mean difference in eating behaviour subscales was analysed using MANCOVA with a Bonferroni-adjusted post-hoc test. Statistically significant was set at p < 0.05.

3. Results

Table 1 shows the socio-demographic characteristics of young adults. Of the 682 respondents, the majority were females (n = 415, 60.9%), aged 18–24 years (n = 536, 78.6%), Indians (n = 321, 47.1%), single (n = 651, 95.5%) and earned less than RM 4000 in a month (n = 649, 95.2%). The mean age of the surveyed young adults was 22.76 ± 2.54 years old, whereas the average income was RM 1317.51 ± 2411.42 in a month during the post-lockdown pandemic phase.

Table 1.

Socio-demographic characteristics of the young adults.

Characteristics	Frequency, n (%)	Mean ± standard deviation
Gender
Male	267 (39.1)	–
Female	415 (60.9)
Age
18–24	536 (78.6)	22.76 ± 2.54
25–30	146 (21.4)
Ethnicity
Malay	253 (37.1)
Chinese	65 (9.5)	–
Indian	321 (47.1)
Others (mixed-race, bumiputra of Sabah and Sarawak)	43 (6.3)
Marital status
Single	651 (95.5)	–
Married	31 (4.5)
Monthly earned income (RM)1
≤ 4000	649 (95.2)	1317.51 ± 2411.42
> 4000	33 (4.8)

¹1 USD = 4.22 Malaysian Ringgit (RM) (as of April 2022).

Table 2 demonstrates changes in the BMI of young adults throughout the COVID-19 pandemic. It is observed that the COVID-19 pandemic led to a statistically significant increase (t = −4.198, p < 0.001) in the BMI of young adults (pre-pandemic BMI = 22.51 ± 5.26 kg/m² versus post-lockdown BMI = 22.87 ± 5.04 kg/m²). In terms of the relative weight change throughout the pandemic, nearly half of the young adults (49.6%, n = 338) had comparable body weight in the post-lockdown as before the pandemic. One-third (32.1%, n = 219) of the young adults put on body weight after the pandemic, with an average relative weight gain of 12.44 ± 9.67%. Conversely, about one-fifth (18.3%, n = 125) of the young adults reported having lighter body weight in the post-lockdown pandemic phase, with an average relative weight loss of 10.10 ± 4.66% (Fig. 1).

Table 2.

Changes in the body mass index of young adults throughout the pandemic.

Body weight status	Frequency, n (%)	Mean ± standard deviation
Pre-pandemic BMI (kg/m2)
Underweight	151 (22.2)
Normal	269 (39.4)	22.51 ± 5.26a
Overweight	92 (13.5)
Obese	170 (24.9)
Post-lockdown BMI (kg/m2)
Underweight	119 (17.4)
Normal	280 (41.1)	22.87 ± 5.04b
Overweight	91 (13.3)
Obese	192 (28.2)

BMI = Body Mass Index.

^a,bSignificant difference was tested using paired samples t-test. Different letters denote statistically significant at p < 0.05.

Fig. 1.

The relative weight change status of young adults throughout the COVID-19 pandemic.

Table 3 indicates the social determinants of eating during the post-lockdown pandemic phase. The findings of Pearson's correlation revealed that ethnicity was positively correlated with cognitive restraint (r = 0.093, p = 0.015), uncontrolled eating (r = 0.090, p = 0.019) and emotional eating (r = 0.093, p = 0.016) in the post-lockdown pandemic phase. On the contrary, monthly earned income was negatively correlated with uncontrolled eating (r = −0.078, p = 0.042) after 22 months of the pandemic. Nevertheless, gender, age, and marital status were not statistically correlated with the eating behaviour of young adults during the post-lockdown pandemic phase.

Table 3.

Social determinants of eating behaviour during the post-lockdown pandemic phase.

Social determinant	Eating behaviour, r (p-value)
	Cognitive restraint	Uncontrolled eating	Emotional eating
Gender	0.029 (0.442)	−0.11 (0.765)	0.037 (0.337)
Age	0.053 (0.165)	−0.03 (0.432)	0.028 (0.469)
Ethnicity1	0.093 (0.015)⁎	0.090 (0.019)⁎	0.093 (0.016)⁎
Marital status	−0.068 (0.078)	−0.046 (0.226)	−0.056 (0.143)
Monthly earned income (RM)	−0.053 (0.169)	−0.078 (0.042)⁎	−0.041 (0.288)

¹Ethnicity was dichotomised into two levels (non-Malay = 0, Malay = 1).

^⁎Significant difference was considered at p < 0.05. Ethnicity and monthly earned income were selected as covariates in the MANCOVA analysis.

Based on the findings of Pearson's correlation, ethnicity and monthly earned income were selected as confounding variables in the MANCOVA analysis. Table 4 compares eating behaviour by relative weight change status of young adults throughout the COVID-19 pandemic. The total standardised score for all eating behaviour subscales was 47.26 ± 19.32 (cognitive restraint), 46.87 ± 21.77 (uncontrolled eating), and 45.76 ± 26.94 (emotional eating). It is noted that there were significant differences in the standardised raw score of cognitive restraint (F = 8.517, p < 0.001), uncontrolled eating (F = 4.458, p = 0.012) and emotional eating (F = 4.909, p = 0.008) across the relative weight change statuses. Young adults in the weight loss category (53.07 ± 20.33) had a significantly higher standardised raw score in cognitive restraints than those in the sustained weight (45.20 ± 19.31, p < 0.001) and weight gain (47.11 ± 18.11, p = 0.013) categories. In addition, young adults in the weight gain category had statistically higher standardised raw scores in uncontrolled eating (49.23 ± 20.84) and emotional eating (48.66 ± 26.41) compared to those in the sustained weight category (uncontrolled eating = 44.42 ± 21.57, p = 0.039 and emotional eating = 42.81 ± 26.30, p = 0.041). On the contrary, there was no significant difference in the standardised raw scores of uncontrolled eating and emotional eating among those in the weight gain and weight loss (uncontrolled eating = 49.33 ± 23.30 and emotional eating = 48.71 ± 28.91) categories.

Table 4.

Comparison of eating behaviour by relative weight change status throughout the COVID-19 pandemic.

Eating Behaviour1	Standardised raw score (Mean ± standard deviation)				F-value	p-value
	Sustained weight (n = 338)	Relative weight loss (n = 125)	Relative weight gain (n = 219)	Total
Cognitive restraint	45.20 ± 19.31a	53.07 ± 20.33b,c	47.11 ± 18.11a,d	47.26 ± 19.32	8.517	<0.001
Uncontrolled eating	44.42 ± 21.57a	49.33 ± 23.30a,c	49.23 ± 20.84b,c	46.87 ± 21.77	4.458	0.012
Emotional eating	42.81 ± 26.30a	48.71 ± 28.91a,c	48.66 ± 26.41b.c	45.76 ± 26.94	4.909	0.008

^a,b,c,dDifferent letters denote statistically significant at p < 0.05 within the same row. The first letter in a row compares the mean difference between sustained weight, relative weight loss, and relative weight change, while the second letter in a row compares the mean difference between relative weight loss and relative weight gain.

¹With the adjustment of ethnicity and monthly earned income (RM). The Bonferroni-adjusted post-hoc test was used to assess within-group variation.

4. Discussion

The National Health and Morbidity Survey 2019 (NHMS 2019) reported that 1 in 2 adults residing in Malaysia is overweight/obese [16]. Findings in the current study indicated that the prevalence of overweight/obese increased marginally from 38.4% (pre-pandemic) to 41.5% (post-lockdown). Likewise, the BMI of young adults rose 1.6%, from 22.51 kg/m² before the pandemic to 22.87 kg/m² after 22 months of the pandemic. A similar pattern of findings was obtained in several studies conducted in Malaysia during the enforcement of MCOs. Overall, young adults in Malaysia were reported to have had a higher BMI during the nationwide lockdowns compared to pre-pandemic [9,10,17].

It is also noteworthy that young adults residing in Malaysia continued to put on body weight throughout the pandemic. Fifteen months into the pandemic, 25.5% of adults in Malaysia have gained more than 5% of their pre-pandemic body weight, according to Tan et al. [14]. The average relative weight gain among those who have gained weight during the COVID-19 pandemic was 9.61% in June 2021 (15 months after the pandemic). When comparing the findings of the current study to those previously reported, it is observed that the relative weight gain of young adults increased to 12.44 ± 9.67% after 22 months of the COVID-19 pandemic in Malaysia. Interestingly, approximately one-fifth (18.3%) of young adults in the current study declared an average relative weight loss of 10.10 ± 4.66% of their pre-pandemic body weight. On the contrary, Tan et al. [14] reported that adults in Malaysia experienced a relative weight loss of 6.65% after 15 months of the pandemic. The relative weight loss among those who have lost weight after 22 months of the pandemic is nearly 1.5 times higher than the previously mentioned study.

Literature has consistently reported that eating behaviour is shaped by individuals' socio-demographic characteristics and economic status [18,19]. Being female, those in their early 20's, those who live with a partner, and those with a wealthier economic status are particularly at heightened risk for disordered eating [[20], [21], [22], [23]]. Despite those previously mentioned, no significant correlations were observed between socio-demographic characteristics and the disordered eating behaviour of young adults throughout the pandemic, except for ethnicity. Ethnicity was positively correlated with cognitive restraint, uncontrolled eating, and emotional eating. These findings go beyond a study by Chan et al. [20], showing that ethnicity was not associated with the risk of eating disorders among Malaysian university students. Besides, the monthly earned income was also reported to be negatively correlated with uncontrolled eating during the COVID-19 pandemic. This finding ties well with a previous study by Haddad et al.[24] in Lebanon, wherein individuals with financial difficulty due to pandemic confinement were reported to display higher levels of eating disorder symptomatology during a pandemic outbreak.

Cognitive restraint refers to constantly and consciously limiting food intake to maintain body weight [25]. The findings of two intervention studies support the notion that cognitive restraint is a predictor of weight loss in overweight/obese populations [26,27]. Cognitive restraint eaters are more likely to lose weight owing to the fact that they are more health-conscious while making food choices [28]. In France, one study pointed out that cognitive restraint eaters tend to choose healthy food items such as green vegetables over French fries and sugar compared to non-cognitive restraint eaters [13]. Emerging results revealed that young adults who lost more than 5% of their pre-pandemic body weight had a significantly higher standardised raw score in cognitive restraint than their counterparts. Interestingly, a similar trend of findings was obtained compared to those previously mentioned literature, even though this study was conducted during the COVID-19 pandemic.

Uncontrolled eating measures the tendency towards overeating due to a loss of control over food intake [13]. Few studies suggested that uncontrolled eating leads to a higher BMI and obesity risk, resulting from a stronger desire for highly palatable foods [[29], [30], [31]]. Moreover, it has been previously reported that food stockpiling is common among households in Malaysia during the enforcement of MCOs [32]. The current findings demonstrated that young adults who gained more than 5% of their pre-pandemic body weight had a significantly higher score in uncontrolled eating than those who managed to sustain weight during the COVID-19 pandemic. Excess food supply in a household, alongside the poor response to food cues, may have offered the opportunity for uncontrolled eating during the pandemic [11].

Emotional eating refers to the tendency towards overeating in the presence of negative emotions such as stress, anxiety, depression, and loneliness [25]. Individuals who eat emotionally often use highly palatable foods to suppress and soothe negative feelings [33]. A recent study by Wong et al. [34] pointed out that the COVID-19 pandemic has worsened the psychological well-being of Malaysian adults. During MCO 1.0, Malaysian adults had higher levels of depression, anxiety, and stress. Tan et al. [35] also indicated that the prevalence of depression during the MCO 1.0 was five times higher than the nationally representative samples in NHMS 2019 [16]. When compared to those who maintained a relatively stable body weight throughout the COVID-19 pandemic, those who gained more than 5% of their pre-pandemic body weight had a significantly higher standardised raw score in emotional eating. Elevated depressive, anxiety, and stress symptoms during the enforcement of home confinement may have emerged as the predisposing factor for emotional eating and weight gain in the aftermath of the COVID-19 pandemic.

Findings in the current study should be interpreted in the context of its limitations. Firstly, this study adopted an online cross-sectional study design in view that face-to-face data collection was not feasible during the COVID-19 pandemic. Due to the nature of the online survey, the survey link can only reach those living in urban/semi-urban regions and with access to the internet at the time of data collection. Secondly, the cross-sectional study design cannot be used to infer the causality between disordered eating behaviour and weight trajectory throughout the COVID-19 pandemic. Thirdly, the self-reported anthropometric measurements may potentially be subjected to recalling bias. Lastly, the reported findings might not be able to be generalised to diverse populations in Malaysia as only a small number of young adults residing in Malaysia responded to this survey. Despite those mentioned earlier, this study is the first to compare young adults' eating behaviour according to their relative weight change status throughout the COVID-19 pandemic in Malaysia.

5. Conclusion and recommendations

Emerging findings suggested that weight trajectory during the COVID-19 pandemic is linked to disordered eating behaviour among young adults in Malaysia. The findings presented in this study can be potentially valuable in formulating weight management strategies in the post-COVID-19 era.

Besides, physical activity levels, duration devoted to sedentary behaviour, food choices, dietary patterns, psychological state, and sleep patterns during the COVID-19 pandemic are also interplaying with the weight trajectory [[8], [9], [10],17,36]. Future studies may investigate the interaction effects of the factors mentioned earlier with weight changes in the aftermath of the COVID-19 pandemic.

Authors' contribution

Seok Tyug Tan: Conceptualization, methodology, supervision, validation, writing- original draft preparation, reviewing, and editing. Thivvyatracyny Mohana Kannan: Data curation and data analysis.

Role of funding sources

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Ethical approval

Ethical approval was granted by the Research Ethics Committee of Management and Science University with the reference number MSU-RMC-02/FR01/12/L1/124.

Declaration of Competing Interest

The authors declare no conflict of interest.