Table 1.
Study characteristics.
| Author, Year, Country | Study Design | Measurement Tool | Participant Characteristics |
Main Findings | Additional Relevant Findings | Study Quality |
|---|---|---|---|---|---|---|
| Aguilar-Martinez et al., 2021 [33], Spain | Cross-sectional (pre-lockdown and during lockdown, 1 cohort) |
Food Frequency Questionnaire (FFQ) (self-reported; online) |
303 high-school students, 14–18 years, 70% female | 2 × healthy marker increase (fruit * and vegetable * consumption). 3 × unhealthy marker decrease (soft drinks *, sweets and pastries *, and convenience foods). Decrease in regularity of meals *. Increase in snacking between meals. Unit of measurement: increase/decrease/no change pre vs. during pandemic |
Reduction in fruit and vegetable consumption, increase in convenience food consumption, decrease in regularly of meals *, and increase in skipping meals * were significantly higher among adolescents from a socioeconomic position perceived to be more disadvantaged *. The highest decrease in the intake of convenience foods was in girls *; the consumption of sweets was the variable that decreased the most in boys *. | High |
| Mastorci et al., 2021 [34], Italy | Cross-sectional (pre-lockdown and during lockdown, 1 cohort) |
KIDMED (self-reported; online) | 1289 school children and adolescents (aged 10–14 yrs), 52% female | 1 × healthy marker increase (adherence to the Mediterranean diet ** (Cohen’s d = −0.157 (s)). Unit of measurement: KIDMED score |
High | |
| Medrano et al., 2021 [35], Spain | Cross-sectional (pre-lockdown and during lockdown, 1 cohort) |
KIDMED (self-reported; online) | 113 school children and adolescents (aged 8–16 yrs), 48% female | 1 × healthy marker increase (adherence to Mediterranean diet *). Unit of measurement: KIDMED score |
High | |
| Munasinghe et al., 2020 [36], Australia | Cross-sectional (1 cohort followed up weekly over 22 weeks, a period spanning before and during lockdown) |
Bespoke questionnaire including food frequency data (online, self-reported) | 582 adolescents (aged 13–19 yrs) from the general population, 61% females |
1 × unhealthy marker decrease (fast food consumption *, Cohen’s d = 0.350 (m)). Unit of measurement: servings per day |
High | |
| James et al., 2021 [37], Wales | Cross-sectional (pre-lockdown and during lockdown, separate cohorts) |
Bespoke questionnaire including food frequency data (self-reported, online) | 2218 (1150 pre-pandemic; 1068 pandemic) school children (aged 8–11 yrs), 51% females | 2 × unhealthy marker decrease (takeaways *, fizzy drink consumption). 1 × unhealthy marker increase (sugary snacks *). 1 × healthy eating marker decrease (fruit/veg consumption). Increase in frequency of breakfast consumption *. Unit of measurement: frequency of consumption per week |
Those who had free school means had significantly lower fruit and veg consumption * and lower sugary snack consumption. | High |
| Kim et al., 2021 [38], Korea | Cross sectional (pre-lockdown and during lockdown, separate cohorts) |
Bespoke questionnaire including food frequency data (self-reported, online) | 105,600 middle- and high-school adolescents (53,461 pre-pandemic; 52,139 pandemic), aged 12–18 yrs, 52% female | 1 × healthy marker decrease (fruit **) 3 × unhealthy marker decrease (fast food **, soda **, sweet drinks **). Higher frequency of eating breakfast **. Unit of measurement: amount per week |
Reporting subjective body shape image as obese was lower in the 2019 group than in the 2020 group **. BMI was slightly though significantly higher in the 2020 group (21.3 vs. 21.5) ** |
High |
| Luszczki et al., 2021 [39], Poland | Cross-sectional (pre-lockdown and during lockdown, separate cohorts) |
Modified Food Frequency Questionnaire (FFQ-6) (parent reported for children, self-reported for adolescents; online) | 1017 (376 pre-pandemic group; 641 during lockdown group) school children (ages 6–12 yrs) and adolescents (aged 13–15 yrs), 51% females | 4 × healthy marker decrease (legumes ** (Cohen’s d = 0.291 (s)), fish * (Cohen’s d = 0.081 (t)), raw veg, fresh fruits). 2 × unhealthy marker decrease (carbonated sugar-sweetened drinks, fast foods ** (Cohen’s d = 0.262 (s)). Decreased snacking * Unit of measurement: frequency of consumption per day or week |
Average | |
| Al Hoirani et al., 2021 [40], Jordan | Cross-sectional (current (during lockdown) and retrospective (pre-lockdown) estimates) | Food Frequency Questionnaire (parent reported for children, self-reported for adolescents; online) | 447 children (6–12 yrs, 51%) and adolescents (13–17 yrs, 49%) from the general population, 52% female | 3 × healthy marker increase (cooked veg for children * (Cohen’s d= −0.08 (t)) and adolescents ** (Cohen’s d = −0.14 (s)), raw veg for children * (Cohen’s d = −0.14 (s)) and adolescents * (Cohen’s d = −0.11 (s)), total fruits for children ** (Cohen’s d = −0.13 (s)) and adolescents ** (Cohen’s d = −0.16 (s)). 7 × unhealthy markers increase (carbonated drinks for children ** (Cohen’s d = −0.15 (s)) and adolescents * (Cohen’s d −0.13 (s)), fries for children * (Cohen’s d = −0.14 (s)) and adolescents ** (Cohen’s d = −0.19(s)), pizza * (Cohen’s d = -0.1(s)) and potato chips ** (Cohen’s d = −0.21(s)) for adolescents, sugar for children * (Cohen’s d = −0.1(s) and adolescents ** (Cohen’s d = −0.19 (s), ice cream for children ** (Cohen’s d= −0.2 (s)) and adolescents ** (Cohen’s d = −0.18 (s)), cake for children ** (Cohen’s d = −0.18 (s), trend in cake for adolescents. Unit of measurement: servings per day |
Increase in BMI for age Z-score **, decrease in thinness and severe thinness ** | Low |
| Androutsos et al., 2021 [41], Greece | Cross-sectional (current (during lockdown) and retrospective (pre-lockdown) estimates) | Bespoke questionnaire including food frequency data (parent-reported; online) | 397 adolescents (12–18 yrs) from the general population, 49% females | 2 × healthy marker increase (fruit ** Cohen’s d = −0.215 (s), vegetables ** Cohen’s d = −0.093 (t)). 3 × unhealthy marker increase (prepacked juices and sodas, salty snacks, sweets ** Cohen’s d = −0.333 (m)). 1 × unhealthy marker decrease (fast food ** (Cohen’s d = 0.045 (t)). Increased snack frequency ** (Cohen’s d = −0.597 (l)). Increased breakfast consumption frequency ** (Cohen’s d = −0.267 (s)). Unit of measurement: servings per day |
Multiple regression analysis showed that body weight increase was associated with increased consumption of breakfast, salty snacks, and total snacks, and with decreased physical activity. | High |
| Horikawa et al., 2021 [42], Japan | Cross-sectional (current (during lockdown) and retrospective (pre-lockdown) estimates) | Bespoke questionnaire including food frequency data (parent-reported; paper postal questionnaire) | 1111 children and adolescents (10–14 yrs) from the general population, 51% females | 2 × healthy marker decrease (fruit ** and veg **). Decreased population of ‘well-balanced dietary intake’ during lockdown **. Unit of measurement: consumption at least twice per day over one month. |
The lower the income group, the greater the rate of decrease in ‘well-balanced dietary intake’ during lockdown. ** | High |
| Kolata et al., 2021 [43], Poland | Cross-sectional (current (during lockdown) and retrospective (pre-lockdown) estimates) | Bespoke questionnaire including food frequency data (self-reported; online) | 1334 school children and adolescents (10–16 yrs), 53% females | 2 × healthy marker increase (fruit ** and veg **). No significant change in markers of unhealthy eating (fast food, fried potato consumption). Unit of measurement: portions per day. |
High | |
| Konstantinou et al., 2021 [44], Cyprus | Cross-sectional (current (during lockdown) and retrospective (pre-lockdown) estimates) | Bespoke food frequency questionnaire (parent-reported; online) | 1509 school children and adolescents (aged 5–14 yrs), 48% females | 1 × healthy marker increase (consumption of fish **). 1 × unhealthy marker increase (consumption of sugary foods **). 1 × unhealthy marker decrease (consumption of ready-made foods). No change for fruit, vegetable, or legume consumption. Increased daily consumption of breakfast *. Unit of measurement: number of days consumed per week. |
High | |
| Muzi et al., 2021 [45], Italy | Cross sectional (one point during lockdown, compared with eating data from one pre-lockdown study) |
Binge Eating Scale (self-reported; online) | 62 adolescents aged 12–17 yrs from the general population; 63% females | No difference in binge eating scale outcomes. Unit of measurement: Binge Eating Scale score |
Pandemic adolescents exhibited more problematic social media usage than their pre-pandemic peers (p < 0.001); more problematic social media usage was correlated with a higher total score of emotional-behavioural symptoms (p < 0.05) (p < 0.05) and more binge-eating attitudes (p < 0.05) (p < 0.05). The role of problematic social media usage was explored as a potential predictor of more total and externalizing problems and binge-eating attitudes, but no predictive model was statistically significant, all p < 0.076 | Average |
| Radwan et al., 2021 [46], Palestine | Cross-sectional (current (during lockdown) and retrospective (pre-lockdown) estimates) | Bespoke questionnaire including food frequency data (parent-reported for <11 yrs, self-reported for 12 + yrs; online) | 6398 school children and adolescents (aged 6–18 yrs, 88% 10–14 yrs), 80% females | 5 × healthy marker increase (overall healthy food rating **, higher median ‘food quality score’ **, increase in students’ rating of their healthy food consumption as very good/excellent ** increased fruit consumption 2+/day **, more home-cooked meals **, and decrease in 1 other (lower veg consumption 2+/day **). 3 × unhealthy marker decrease (decreased fast food consumption **, decreased consumption of sodas/sweet tea 3+/day **, decreased desserts/sweets 4+ times/week **). Decreased snacking **. Decrease in food quantity ** Units of measurement: number of days consumed on per week |
Boys had higher food quality scores during the pandemic **, whereas girls had higher pre-pandemic scores in food quality. Students aged 6–9 years exhibited a higher food quality score during the COVID-19 period than before the COVID-19 period **, whereas students aged 10–14 ** and 15–18 ** years attained a lower score during the COVID-19 period. Students aged 6–9 and 15–18 years had higher food quantity scores during the COVID-19 period **, but students aged 10–14 years had a lower score. Decrease in the proportion of students whose families bought groceries every day **. Increase in students who agreed or strongly agreed with the idea of choosing food according to calorie content and healthy properties when they buy **. Marked increase in students reporting fears about food hygiene from outside during the pandemic **. |
High |
| Segre et al., 2021 [47], Italy | Cross-sectional (one time point during data, change data) |
Bespoke (self- and parent-reported, video interview) | 82 school children and adolescents (aged 6–14 yrs), 54.9 % primary school, 45.1 % middle school, 46% females | A significant proportion of children reported a perceived change in their eating habits *. Non-significant trend towards children reporting eating more during the pandemic, with a non-significant increase in the consumption of junk food, snacks, and sweets. Units of measurement: increase/decrease /no change pre vs. during pandemic |
‘Changed eating behaviours’ were more significant in primary school children * compared with middle school children. Anxiety levels were not found to be significantly associated with changes in eating behaviours. Higher frequency of mood symptoms was associated with changes in dietary habits *. | High |
| Yu et al., 2021 [48], China | Cross-sectional (current (after lockdown) and retrospective (pre-lockdown) estimates) | Bespoke questionnaire including food frequency data (self-reported; online) | 2824 high school students, 72% females | 3 × healthy eating marker increase (fish * Cohen’s d = −0.077 (t), fresh veg * Cohen’s d = −0.045 (t), preserved veg * Cohen’s d = −0.056 (t)). Units of measurement: frequency of consumption per week |
Low |
* p < 0.05, ** p < 0.01. (t) = trivial effect size, (s) = small effect size, (m) = moderate effect size, (l) = large effect size.