TABLE 1.
Characteristics of quantitative studies
| Author(s) (year) | Country | Participants | Methods | Key findings | Study quality | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| N (% female) a | Age M (SD) | Race and ethnicity | Sexual orientation | Socioeconomic status | Diagnosis | Design | Measures | ||||
| Akgül et al. (2021) | Turkey | 38 (95%) | 15.1 (1.6) | Not reported | Not reported | Not reported | AN‐R n = 26; AN‐BP n = 5; AAN n = 3; BN n = 3; OSFED n = 1 | Cross‐sectional | ED examination; depression; anxiety; obsessive–compulsive symptoms | 42% reported improved ED symptomatology, 37% reported no change, 21% reported worse ED symptomatology from pre‐ to during lockdown; 24% reported that lockdown affected access to ED‐related healthcare; depression score had the highest predictive value for ED behavior (r 2 = .537) | 3 |
| Baceviciene and Jankauskiene (2021) | Lithuania | 230 (79%) | 23.9 (5.4) | Not reported | Not reported | Not reported | n/a | Longitudinal | Attitudes toward appearance; BI; ED examination; self‐esteem | No change in DE or self‐esteem from pre‐ to during lockdown; body appearance evaluation (women only), media pressures (women only), and internalization of thin/low body fat appearance standards (men: d = 1.46; women: d = 1.18) increased from pre‐ to during lockdown | 3 |
| Baenas et al. (2020) | Spain | 74 (96%) | 32.1 (12.8) | Not reported | Not reported | Not reported | AN n = 19; BN n = 12; BED n = 10; OSFED n = 33 | Cross‐sectional | ED inventory; food addiction; symptom checklist; temperament and character; telephone survey | 26% reported worsened symptoms during confinement (highest in AN and OSFED), 74% reported improvements or no change; patients with worsened symptoms reported lower self‐directedness (d = 0.51) and higher prevalence of future concerns (d = 0.51), nonadaptive reactions (d = 0.79), symptoms of anxiety (d = 0.89) and depression (d = 0.96), adverse situations (d = 0.62), and familiar conflict (d = 0.68) | 2 |
| Bellapigna et al. (2021) | United States | 239 (gender: 79% women; 1.3% nonbinary/nonconforming) | 24.7 (11.1) | 6.7% Black/African American; 10% Hispanic; 5.9% Asian; 67.8% White; 9.2% multiracial/biracial; 0.4% “other” | Not reported | Education: 25.6% high school degree or equivalent; 46.2% some college; 10.1% 2‐year degree; 11.8% bachelor's degree; 3.4% master's degree; 2.1% doctorate; 2.8% other | 6.3% diagnosed with ED | Cross‐sectional | Need for structure; loneliness; social networking; body appreciation; eating attitudes; social phobia; patient health | 64% reported more disturbances in BI during COVID; loneliness (β = −.138), negative BI (β = .253), and social media exposure (β = −.161) predicted DE; loneliness (β = −.485), negative BI (β = .123), and social media exposure (β = −.154) predicted depressive symptoms | 3 |
| Branley‐Bell and Talbot (2020) b | United Kingdom | 129 (94%) | 29.3 (9.0) | Not reported | Not reported | Not reported | 62% current ED/relapse; in recovery 6.2% <3 m, 6.2% 3–12 m, 25.6% >12 m | Cross‐sectional | Mental well‐being; perceived stress; social support; sense of control; rumination | 87% reported worsened symptoms as a result of COVID, 30% reported symptoms were much worse, 2% reported slight improvement, 9% reported no change; changes in living situation, social isolation, usual support network(s), physical activity, and time spent online impacted ED symptoms | 3 |
| Branley‐Bell and Talbot (2021) b | United Kingdom | 58 (98%) | 30.9 (11.1) | Not reported | Not reported | Not reported | 63.8% current ED/relapse; 36.2% in recovery; AN n = 28; BN n = 7; OSFED n = 3; BED n = 2; symptoms of multiple EDs n = 12; undisclosed ED n = 7 | Longitudinal | Mental well‐being; perceived stress; social support; sense of control; rumination | 15.5% reported relapsing, 19% reported recovering, and 65.5% reported no change in ED status from during to postlockdown; higher perceived control associated with recovery | 2 |
| Breiner et al. (2021) | United States | 159 (91%) | 27.6 (11.7) | 90.6% White; 5% Hispanic/Latino; 6.3% Asian; 0.6% American Indian or Alaska Native; 0.6% Native Hawaiian or Pacific Islander; 1.3% Native American | 74.8% heterosexual; 2.5% homosexual; 14.5% bisexual | Education: 5% high school graduate; 8.8% less than 2 years of college; 0.6% technical or vocational program; 6.9% associate degree; 62.3% college graduate; 14.5% master's degree; 1.9% doctorate | AN n = 22; BN n = 8; BED n = 4; “other” n = 3 | Cross‐sectional (retrospective) | ED examination; exercise behaviors; exercise motives; ED diagnosis | No significant changes in exercise or eating pathology from pre‐ to during COVID; eating pathology increased in participants with a prior ED diagnosis (d = 0.26), but decreased in participants without a prior ED diagnosis (d = −0.14); there were decreases in episodes of eating more than usual (d = −0.23) and loss of control over eating (d = −0.23), but no changes in objective binge eating (d = −0.04), self‐induced vomiting (d = −0.05), or laxative use (d = −0.14); participants reported increased endorsement of pressure to get in shape from pre‐ to during the pandemic | 3 |
| Buckley et al. (2021) b | Multiple | 204 (86%) | 27.0 (8.1) | Not reported | Not reported | Not reported | 10.7% current ED; 32.8% previous diagnosis (AN n = 29; BN n = 11; ON n = 9; BED n = 7; “other” n = 11) | Cross‐sectional | Eating attitudes; BI and food relationship | 34.8% reported worse BI, 50.5% reported no change, and 14.6% reported better BI from pre‐ to during COVID; 32.8% reported worse relationship with food, 53.0% reported no change, and 14.1% reported better relationship with food from pre‐ to during COVID | 3 |
| Castellini et al. (2020) | Italy | Patients/healthy controls: 74/97 (sex assigned at birth: 100/100% female) | Patients/healthy controls: 31.7/30.5 (12.8/10.9) | 100% White | Not reported | Not reported | AN n = 37; BN n = 37 | Longitudinal | Brief symptom inventory; ED examination; psychological distress | Patients reported increased compensatory exercise during lockdown (AN: d = 0.32; BN: d = 0.30); patients with BN/previously remitted patients reported increased binge eating after lockdown (d = 0.32); household arguments (d = 0.62) and fear for safety of loved ones (d = 0.67) predicted increased symptoms; patients with BN reported more severe COVID‐related posttraumatic symptoms than patients with AN and healthy controls, predicted by childhood trauma (β = .34) and insecure attachment (β = .57) | 1 |
| Cecchetto et al. (2021) | Italy | 365 (73%) | 35.1 (13.6) | Not reported | Not reported | Not reported | n/a | Cross‐sectional | Binge eating screener; alexithymia; anxiety; eating behaviors; patient health; perceived stress | Binge eating and emotional eating decreased from during to postlockdown; emotional eating was predicted by higher depression, anxiety, lower quality of personal relationships, and lower quality of life; increase in binge eating was predicted by higher stress | 2 |
| Chan and Chiu (2021) | Hong Kong | 316 (71%) | 25.1 (5.0) | Not reported | Not reported | 88.3% university educated | n/a | Cross‐sectional | ED screening; patient health; anxiety; psychological well‐being; eating behaviors and emotions | Among individuals with elevated depression, those who attributed depression to COVID reported higher levels of symptoms; no effect on anxiety | 3 |
| Christensen, Forbush, et al. (2021) | United States | 579 (gender identity: 76% women; 2% “other”) | 21.8 (5.3) | 9% Hispanic; 91% non‐Hispanic; 84.1% White; 3.5% Black/African American; 1% American Indian/Alaskan Native; 5.5% Asian/Pacific Islander; 5.2% multiracial; 0.7% not disclosed | Not reported | Years of posthigh school: 20.3% <1 year; 18.7% 1 year; 19.4% 2 years; 22.0% 3 years; 9.9% 4 years; 2.4% 5 years; 5.0% 6+ years 5%; 2.3% continuing education; food insecurity: 52.8% none; 6.6% household food insecurity; 40.6% individual food insecurity | AN n = 4; BN n = 75; BED n = 14; OSFED n = 135; no ED diagnosis n = 344 | Cross‐sectional | Clinical impairment; ED diagnosis; food insecurity | Students with food insecurity showed higher prevalence of ED diagnoses and reported greater frequency of objective binge eating, compensatory fasting, and ED‐related impairment compared with individuals without food insecurity; there were no differences in food insecurity before or during the beginning of the COVID pandemic; participants who identified as Black were significantly more likely to report individual food insecurity relative to other racialized groups | 2 |
| Conceição et al. (2021) | Portugal | COVID/non‐COVID group: 35/66 (94/83%) | COVID/non‐COVID group: 50.8/50.1 (12.4/10.7) | Not reported | Not reported | Education (COVID/non‐COVID group): ≤6 years (42.9/31.8%; 9–12 years (34.3/48.5%); college degree (22.9/19.7%); professional status: student (2.9/1.5%); employed (62.9/57.6%); unemployed (11.4/25.8%); retired: (22.9/15.2%) | n/a | Cohort | ED examination; repetitive eating; depression, anxiety, stress; impulsivity | Participants assessed post‐COVID showed a greater increase in weight concern scores (ƞ 2 p = 0.094) and repetitive eating (ƞ 2 p = 0.076) compared with participants assessed pre‐COVID; no difference between groups in shape concern, food concerns, or restraint eating | 1 |
| Coulthard et al. (2021) | United Kingdom | 620 (88%) | 39.9 (14.0) | 88% White‐British/European; 6% Asian/British Asian; 1% Black/Black British; 4% “other”; 1% not disclosed | Not reported | Occupation: 39% professional; 21% intermediate; 15% manual; 25% not working; food insecurity: 65% none; 29% mild; 6% moderate/severe | n/a | Cross‐sectional | Food consumption; ED symptoms; coping strategies; anxiety; food insecurity | Emotional eating decreased from pre‐ to during lockdown; women and those isolating at home were more likely to report higher emotional eating during lockdown; there was no differences in eating behavior based on occupation or ethnicity; higher emotional eating during lockdown associated with higher BMI, higher prelockdown emotional eating, and maladaptive coping strategies | 2 |
| Czepczor‐Bernat et al. (2021) | Poland | 671 (100%) | 32.5 (11.4) | 98.8% White; 0.3% mixed race; 0.9% “other” | 91.3% heterosexual; 1.6% lesbian; 4.9% bisexual; 0.7% pansexual/queer; 0.9% asexual; 0.6% “other” | Education: cluster 1 32% secondary/technical school; cluster 2 32% secondary/technical school; cluster 3 37% master's degree; cluster 4 45% master's degree | n/a | Cross‐sectional | COVID‐related stress; COVID‐related anxiety; ED inventory; BI | Higher levels of ED symptoms and negative BI were observed in women with excess body weight, high anxiety, and stress related to COVID as compared with women with a healthy body weight and with low levels of anxiety and stress | 3 |
| Félix et al. (2021) | Portugal | 24 (100%) | 50.9 (12.8) | Not reported | Not reported | Education: 29.2% elementary school; 25.0% middle school; 12.5% high school; 33.3% college degree; employment: 66.7% employed; 8.3% unemployed; 25.0% retired | n/a | Cross‐sectional | Depression, anxiety, stress; impulsivity; DERS; ED examination; loss of control over eating; repetitive eating; ED symptoms; impact on emotions, eating, and weight; COVID impact | Living with fewer people, higher difficulties in dealing with emotionally activating situations, and higher fear of gaining weight during lockdown associated with greater fear of gaining weight, greater fear of losing control over eating, and greater DE psychopathology | 2 |
| Fernández‐Aranda, Munguía, et al. (2020) | Spain | 121 (86%) | 33.7 (15.8) | Not reported | Not reported | Not reported | AN n = 55; BN n = 18; OSFED n = 14 | Cross‐sectional | COVID isolation eating scale | Patients with AN reported a positive response to treatment during confinement; no significant changes found in patients with BN; patients with OSFED reported an increase in eating symptomatology and psychopathology; patients with AN reported greatest dissatisfaction and accommodation difficulty with remote therapy | 3 |
| Fernández‐Aranda, Casas, et al. (2020) Study 1 | Spain | 32 (91%) | 29.2 (range 16–49 years) | Not reported | Not reported | Not reported | AN n = 13; BN n = 10; BED n = 4; OSFED n = 5 | Cross‐sectional | Telephone survey | Most patients presented worries about increased uncertainties, such as the risk of COVID infection of themselves or their loved ones, the negative impact on their work, and their treatment; 38% reported impairments in ED symptomatology; 56% reported additional anxiety symptoms | 3 |
| Flaudias et al. (2020) | France | 5738 (75%) | 21.2 (4.5) | Not reported | Not reported | University students, 48.8% with scholarship | 38.3% at risk for ED symptoms | Cross‐sectional | Anxiety and depression; perceived stress; ED inventory; ED screening; ideal body stereotypes | Greater likelihood of binge eating (BE) and dietary restriction (DR) over past week and/or future intentions to binge eat (FBE) and restrict (FDR) associated with lockdown‐related stress (all odds ratios [OR]; BE = 1.12; DR = 1.17; FBE = 1.33; FDR = 1.12), exposure to COVID‐related media (BE = 1.02; DR = 1.05; FBE = 1.20; FDR = 1), female gender (BE = 1.40; DR = 1.79; FBE = 1.09; FDR = 1.48), greater levels of anxiety (BE = 1.09; DR = 1.11; FBE = 0.95; FDR = 1.04) and depression (BE = 1.14; DR = 0.94; FBE = 1.40; FDR = .95), low impulse regulation (BE = 1.10; DR = 1.04; FBE = 1.23; FDR = 1.12), higher BMI (BE = 1.26; DR = 1.07; FBE = 1.19; FDR = 1.04), body dissatisfaction (BE = 1.08; DR = 1.80; FBE = 0.84; FDR = 2.05), and concurrent probable ED (BE = 2.82; DR = 2.65; FBE = 2.11; FDR = 2.58) | 3 |
| Giel et al. (2021) | Germany | 42 (81%) | 45.5 (12.6) | Not reported | Not reported | Not reported | BED n = 17 | Longitudinal | ED examination; perceived stress; binge eating frequency; depression; emotion regulation; coherence | Binge eating frequency (χ 2 = 15.22), general ED pathology (χ 2 = 35.52), and depressive symptoms (χ 2 = 5.41) increased from pre‐ to post‐COVID; individuals scoring high on reappraisal and sense of coherence scored lower on general ED pathology | 3 |
| Graell et al. (2020) | Spain | Day hospital/outpatient clinic: 27/338 (93/87%) | Day hospital/outpatient clinic: 13.2/14.7 (3.0/2.3) | Not reported | Not reported | Not reported | ARFID n = 48; AN n = 255; BN n = 26; OSFED n = 37 | Cross‐sectional (retrospective) | Outpatient and face‐to‐face consultations | 42% reported reactivation of ED symptoms following COVID confinement (>adolescents); 68% of patients and their families reported onset of confinement and 41% reported social isolation from peers as influencing factors for admission | 2 |
| Gullo and Walker (2021) | United States | 143 (gender: 53% ciswomen; 1% transmen; 0% transwomen; 1% nonbinary) | 77% 18–44 years | 79.7% White: 9.8% Black: 4.2% Asian; 3.5% Latinx: 1.4% mixed; 0.7% Native American; 0.7% “other” | Not reported | Household annual income (USD): 2.8% <15,000; 9.1% 15,000–29,999; 11.9% 30,000–49,999; 23.8% 50,000–75,999; 16.8% 75,000–99,999; 17.5% 100,000–150,000; 14.0% >150,000; 4.2% prefer not to answer | n/a | Cross‐sectional (retrospective) | Depression, anxiety, stress; BI; binge eating; body satisfaction | Time spent videoconferencing increased from pre‐ to post‐COVID; appearance dissatisfaction increased (β = .75), but appearance orientation decreased (β = .75) following lockdown; no change in binge eating from pre‐ to postlockdown; videoconferencing time did not predict BI or binge eating postlockdown | 3 |
| Haddad et al. (2020) | Lebanon | 407 (51%) | 30.6 (10.1) | Not reported | Not reported | Education: 90.9% university level; 9.1% secondary school or lower | n/a | Cross‐sectional | Boredom; ED examination; quarantine/ confinement stressors; fear of COVID; anxiety; physical activity | Dietary restraint (DR), eating concerns (EC), shape concerns (SC) and weight concerns (WC) were associated with female gender (all β; EC = 0.52; SC = 0.19; WC = 0.20), higher anxiety (EC = 0.04; SC = 0.23; WC = 0.19), sense of insecurity (EC = 0.41), greater fear of COVID (DR = 0.02; SC = 0.20; WC = 0.12), higher BMI (DR = 0.05; EC = 0.06; SC = 0.39; WC = 0.41), physical activity (DR = 1.04; EC = 0.43; SC = 0.15; WC = 0.19), and a higher number of adults living together in quarantine/confinement (SC = 0.10; WC = 0.15) | 2 |
| Haddad et al. (2021) | Lebanon | 407 (51%) | 30.6 (10.1) | Not reported | Not reported | Household monthly income (USD): 31.0% no income; 20.4% <1000; 29.1% 1000–2000; 19.3% >2000 | n/a | Cross‐sectional | Boredom; ED examination; quarantine/ confinement stressors; fear of COVID; anxiety; physical activity; perceived weight change | Longer confinement duration (OR = 1.07), higher anxiety (OR = 1.05), and higher eating concerns (OR = 1.81) associated with higher weight change perception; greater fear of COVID (OR = 0.96) and higher self‐reported weight change (OR = 0.47) associated with lower weight change perception | 2 |
| Jordan et al. (2021) | United States | 140 (89%) | 39.8 (6.9) | 88.4% White | Not reported | 82.2% middle to upper‐middle class | n/a | Cross‐sectional | Perceived stress; concern about weight gain; ED examination; emotional eating |
Disordered eating associated with concern about weight gain before (β=.18) and during (β=.32) COVID; stress and concern about weight gain during COVID predicted variance in eating pathology among caregivers (r 2 = 0.48). |
2 |
| Keel et al. (2020) | United States | 90 (88%) | 19.5 (1.3) | 22% Latinx; 78% White; 12% Black/African American; 4% Asian; 1% American; Indian/Alaskan Native; 3% “other” | 89% heterosexual | Not reported | n/a | Longitudinal | Weight perception; physical activity; eating; concerns about weight and shape; body, eating, exercise comparisons; ED diagnosis; screen time | Participants reported increased body weight (d = 0.23), eating (d = 0.54), screen time (d = 1.08), and concerns about weight/shape (d = 0.93) and eating (d = 0.79), and decreased physical activity (d = −0.63) from pre‐ to post‐COVID; no change in weight or BMI, but participants reported shifts in body weight perception from pre‐ to post‐COVID | 3 |
| H. Kim, Rackoff, et al. (2021) | United States | Pre‐/during COVID: 3643/4970 (sex assigned at birth: 73/70% female; 0.05/0.02% intersex; gender identity: 70/68% women; 4/2% trans, nonconforming, or self‐identify) | Not reported | Pre‐/during COVID: 12/10% Hispanic; 88/90% non‐Hispanic; 72/75% White; 9/6% Black/African American; 10/14% Asian; 0.7/0.5% American Indian or Alaskan Native; 0.4/0.2% Native Hawaiian or Pacific Islander; 8/5% multiracial | Pre‐/during COVID: 71/81% heterosexual; 29/19% lesbian, gay, bisexual, queer, questioning, or self‐identify | Not reported | Pre‐/during COVID: AN n = 64/88; BN/BED n = 320/643 | Longitudinal | Anxiety; posttraumatic stress; patient health; presence of EDs; insomnia; alcohol use | Depression (χ 2 = 21.67), alcohol use disorder (χ 2 = 67.26), BN/BED (χ 2 = 20.83), and comorbidity (χ 2 = 6.83) were greater during than before COVID; posttraumatic stress disorder was lower during than pre‐COVID (χ 2 = 5.46); no differences in anxiety, insomnia, AN, or suicidality between pre‐ and during COVID; no effect of gender, ethnicity/racialized group, or sexuality on EDs | 2 |
| S. Kim, Wang, et al. (2021) | United States | 7317 (59%) | 50.6 (16.1) | 64.7% non‐Hispanic White; 7.9% non‐Hispanic Black; 16.8% Hispanic; 5.1% non‐Hispanic Asian; 0.9% Native American; 4.5% “other”; 0.2% not disclosed | Not reported | Education: 5.4% <high school; 16.7% high school or less; 22.8% some college; 14.3% associate degree; 24.3% bachelor's degree; 16.5% advanced college degree | Diagnosed ED n = 157; unsure about ED status n = 122 | Longitudinal | Patient health; perceived stress; loneliness | Individuals with EDs/unsure EDs reported higher levels of psychological distress (all B; EDs = 2.18; unsure EDs = 2.01), stress (EDs = 1.17; unsure EDs = 2.08), and loneliness (unsure EDs = 0.90) compared to those without EDs; those unsure about their EDs reported initial decreases in stress and loneliness, but started increasing again since institution of virus containment procedures; levels of loneliness among those with EDs increased initially then began to decrease; individuals with EDs showed steady decreases in stress; identifying as Black, older age, and higher education associated with lower psychological distress (Black = −0.59; age = −0.03; education = −0.03), stress (age = −0.04; education = −0.13), and loneliness (Black = −0.10; age = −0.01); female gender and identifying as Asian associated with higher psychological distress (female = 0.61; Asian = 0.29), stress (female = 0.56; Asian = 1.07), and loneliness (female = 0.14) | 1 |
| Koenig et al. (2021) | Germany | Pre/postlockdown: 324/324 (69/69%) | 14.9 (1.9) | Not reported | Not reported | Family affluence (pre‐/postlockdown): low (1.9/1.9%); medium (24.4/ 21.6%); high (73.9/76.5%) | n/a | Longitudinal | Strengths and difficulties; patient health; weight concerns; ED examination; quality of life; suicidality | No differences between pre‐ and postlockdown samples in emotional and behavioral problems, depression, thoughts of suicide/suicide attempts, ED symptoms, or quality of life | 1 |
| Larkin (2021) | United States | 290 (not reported) | Range 18–25 years | 0.3% American Indian/Alaskan Native; 2.8% Asian; 10.3% Black/African American; 14.5% Hispanic/Latino; 1.7% biracial/multiracial; 80.7% White | Not reported | Not reported | n/a | Cross‐sectional (retrospective) | Physical activity, social media use; subjective well‐being; BI | 32.7% increase in negative BI perceptions from pre‐ to post‐COVID | 3 |
| Leenaerts et al. (2021) | Belgium | 15 (100%) | Median (IQR) = 23 years (21.5–25.5) | 87% European; 13% Asian | Not reported | Not reported | BN | Longitudinal | Affect; location; social context; binge eating frequency | Patients reported higher negative affect (β = .15), lower positive affect (β = −.10), and changes in surroundings and social context (at home: β = 3.19; with housemates: β = 3.91; with friends: β = −2.45; with family: β = .99; with partner: β = −2.39) from pre‐ to postlockdown; changes of negative affect associated with binge eating frequency during lockdown (β = .61) | 2 |
| Lessard and Puhl (2021) | United States | 452 (gender identity: 55% women; 2% transgender; 1% “other”; sex assigned at birth: not reported) | 14.9 (2.1) | 69.9% White; 8.2% Black/African American; 8.0% Latinx; 6.6% multiethnic; 5.5% Asian/Pacific Islander; 1.8% “other” | Not reported | 72% had a parent with a college degree or higher | n/a | Cross‐sectional | Body dissatisfaction; exposure to weight stigma on social media; experienced weight stigma | 53% reported increased exposure to weight stigmatizing social media content; 41% reported increased body dissatisfaction from pre‐ to post‐COVID (>girls; higher weight), 49% reported no change, 10% reported a decrease | 3 |
| Lin et al. (2021) | United States | Not reported | Range 8–26 years | Not reported | Not reported | Not reported | Patients with any ED diagnosis | Longitudinal | COVID‐related trends in ED care‐seeking | Inpatient admissions, hospital bed‐days, outpatient inquiries increased over time post‐COVID compared to stable volume pre‐COVID; outpatient assessments decreased initially following COVID‐related limitations, then rebounded | 2 |
| Machado et al. (2020) | Portugal | 43 (95%) | 27.6 (8.5) | Not reported | Not reported | Not reported | AN n = 20; BN n = 14; BED n = 2; OSFED n = 7 | Longitudinal | ED examination; clinical impairment; impulsivity; difficulties in emotion regulation; COVID impact | Of 26 patients in treatment 31% remained unchanged, 27% deteriorated, and 42% improved; of 17 participants not in treatment 53% remained unchanged, 18% deteriorated, and 29% improved from during to postlockdown; higher impact correlated with ED symptoms, impulsivity (r = .380), psychopathology (r = .451), emotion regulation difficulties (r = .393) and clinical impairment (r = .569) | 3 |
|
Martínez‐de‐Quel et al. (2021) |
Spain | 161 (37%) | 35.0 (11.2) | Not reported | Not reported | Not reported | n/a | Longitudinal | Eating attitudes | No change in ED risk from before to during COVID lockdown | 2 |
| Meda et al. (2021) | Italy | 358 (80%) | 21.3 (2.1) | Not reported | Not reported | Not reported | Not reported | Longitudinal | ED inventory; eating habits; obsessive–compulsive symptoms; anxiety; depression | Only students with ED history reported an increase in ED symptomatology from pre‐ to postlockdown (β = .1). | 3 |
| Monteleone, Cascino, Marciello, et al. (2021) | Italy | 312 (gender identity: 96% women; 0.3% nonbinary) | AN: 26.9 (10.3); other EDs: 32.3 (13.5) | Not reported | Not reported | Employment (AN/other EDs): paid job (26/39%); student (56/49%) | AN n = 179; BN n = 63; BED n = 48; OSFED n = 22 | Cross‐sectional (retrospective) | Factors related to COVID concerns; illness duration; treatment‐related variables; ED psychopathology | General (GP) and specific psychopathology (SP) worsened from pre‐ to postlockdown; perceived low quality of therapeutic relationships (GP: β = −.16; SP: β = −.22), fear of contagion and increased isolation (GP: β = .22; SP: β = .22), reduced satisfaction with relationships (SP: β = .24), and reduced social support (GP: β = .23) associated with worsened psychopathology; no effect of intimate relationships, illness duration, diagnosis, economic change, or type of treatment | 2 |
| Monteleone, Marciello, et al. (2021) | Italy | 312 (gender identity: 96% women; 0.3% nonbinary) | 29.2 (12.1) | Not reported | Not reported | Not reported | AN n = 179; BN n = 63; BED n = 48; OSFED n = 22 | Cross‐sectional (retrospective) | Anxiety; posttraumatic stress; obsessive–compulsive symptoms; patient health; ED inventory | General and specific psychopathology worsened from pre‐ to postlockdown; symptoms persisted postlockdown, apart from suicide ideation; individuals with AN reported higher anxiety, obsessive–compulsive symptoms, suicide ideation, and physical activity levels, and lower binge eating; no effect of age or illness duration | 2 |
| Nisticò et al. (2021) | Italy | 40 (95%) | 30.9 (14.2) | 100% White | Not reported | Not reported | AN n = 15; BN n = 11; BED n = 14 | Longitudinal | ED examination; depression, anxiety, stress; psychological distress | Posttraumatic stress (IES‐R total score: η 2 p = 0.145) and ED symptoms (η 2 p = 0.142–0.249) improved from during to postlockdown; no change in stress, anxiety, or depression | 2 |
| Pfund et al. (2020) | United States | 438 (gender identity: 100% women; 434 ciswomen and 4 transwomen) | 31.3 (12.7) | 11% African American/Black; 22% Asian/Asian American; 52% European American/White; 10% Latinx American/Hispanic; 2% Middle Eastern, and 3% “other”/not disclosed | Not reported | Not reported | n/a | Cross‐sectional | Body surveillance; appearance comparison; body satisfaction | Time video chatting increased from pre‐ to post‐COVID (d = 0.53). Time video chatting not associated with appearance satisfaction; self‐objectification moderated relationship between time video chatting and appearance satisfaction (B = −.04 for face satisfaction and B = −.02 for body satisfaction); participants who spent more time engaged with their families over video chatting services reported greater face (r = .21) and body (r = .17) satisfaction | 2 |
| Phelan et al. (2021) | Ireland | 1031 (sex assigned at birth: 100% female) | 36.7 (6.6) | 97% White; 2% Asian; 0.3% Black; 0.5% “other” | Not reported | Not reported | n/a | Cross‐sectional | Mental health symptoms, diet, exercise | Participants reported an increase in binge eating from pre‐ to during COVID | 3 |
| Philippe et al. (2021) | France | 498 (52%) | 7.3 (2.3) | Not reported | Not reported | Not reported | n/a | Cross‐sectional (retrospective) | Eating difficulties; eating behavior; parental feeding practices | Parents reported an increase in their child's emotional overeating, food responsiveness, food enjoyment, and appetite, but no change in their child's pickiness from pre‐ to during lockdown; boredom predicted increased food responsiveness (β = .14), emotional overeating (β = .20), and snack frequency between meals (β = .28) | 2 |
| Phillipou et al. (2020) | Australia | 5469 (96% women; 3% preferred to self‐describe) | 30.5 (8.2) | Not reported | Not reported | Not reported | AN n = 88; BN n = 23; BED n = 6; OSFED n = 4; UFED n = 68; recovering/in recovery n = 10 | Cross‐sectional | Depression, anxiety, stress; ED examination | Participants with ED history reported increased restricting (64.5%), binge eating (35.5%), purging (18.9%), and exercise behaviors (47.3%); participants without ED history reported both increased restricting (27.6%) and binge eating behaviors (34.6%), but decreased exercise (43.4%) from pre‐ to during COVID | 3 |
| Pikoos et al. (2020) | Australia | 216 (gender: 88% women; 0.01% “other”) | 32.5 (11.8) | Not reported | Not reported | Not reported | n/a | Cross‐sectional | Dysmorphic concern; depression, anxiety, stress; appearance‐focused behaviors | Appearance‐focused behaviors decreased in participants with low dysmorphic concerns, but not in participants with high dysmorphic concerns from pre‐ to during COVID; living alone, younger age, higher dysmorphic concern, and greater distress over beauty service closure predicted appearance‐focused behaviors | 2 |
| Puhl et al. (2020) | United States | 584 (gender identity: 64% women; 1% “other”) | 24.6 (2.0) | 30.2% White; 16.8% African American/ Black; 17.1% Hispanic; 24.3% Asian American; 11.6% “other” | Not reported | 31.5% lower class; 20.0% lower middle class; 17.4% middle class; 19.4% upper middle class; 11.7% upper class (assumed self‐report) | n/a | Longitudinal | Psychological distress; eating behaviors; binge eating; physical activity; weight stigma | Pre‐COVID experiences of weight stigma predicted higher levels of depressive symptoms (β = .15), stress (β = .15), eating as a coping strategy (β = .16), and an increased likelihood of binge eating during COVID (OR = 2.88), but were unrelated to physical activity; no effect of gender | 2 |
| Ramalho et al. (2021) | Portugal | 254 (83%) | 35.8 (11.8) | Not reported | Not reported | Education: 13.0% high school; 27.2% bachelor's degree; 59.8% master's degree/doctorate | n/a | Cross‐sectional | DE behaviors; COVID impact; depression, anxiety, stress; ED symptoms | Psychosocial impact of COVID predicted DE behaviors mediated through psychological distress (>women; younger age) (β = .10); psychosocial impact of COVID associated with emotional eating (r = .23) and uncontrolled eating (r = .18) | 2 |
| Richardson et al. (2020) b | Canada | 439 (gender: 80% women; 2% transgender; 11% did not disclose) | Not reported | Not reported | Not reported | Not reported | AN n = 83; BN n = 44; ARFID n = 4; BED n = 66; OSFED n = 9; undisclosed/ undiagnosed n = 233 | Cross‐sectional (retrospective) | Telephone survey | Service utilization, ED symptoms, anxiety, and depression increased from pre‐ to during COVID among patients with EDs | 3 |
| Robertson et al. (2021) | United Kingdom | 264 (78%) | Range 18–79 years | 92% White | Not reported | Not reported | 13.8% current or past ED diagnosis | Cross‐sectional | Perceived change in eating, exercise, and BI; patient health | 53% reported more difficulty regulating eating; 60% reported more preoccupation with food/eating; 50% reported exercising more; 68% reported thinking more about exercise; 49% reported more appearance concerns from pre‐ to during lockdown (>women; participants with past/current ED); psychological distress was correlated with finding it more difficult to control/regulate one's eating (rs = .36), being more preoccupied with food/eating (rs = .29), thinking more about exercise (rs = .17), and being more concerned about one's appearance (rs = .41) | 3 |
| Scharmer et al. (2020) | United States | 295 (65%) | 19.7 (2.0) | 48% White; 21% African American; 11% Asian; 14% Hispanic/Latino; 1% Native American; 1% Native Hawaiian/Pacific Islander; 3% “other” | Not reported | Not reported | n/a | Cross‐sectional | ED examination; compulsive exercise; anxiety; fear of illness and virus evaluation; intolerance of uncertainty; physical activity | COVID anxiety and intolerance of uncertainty was associated with ED pathology, but not compulsive exercise; trait and COVID intolerance of uncertainty moderated associations between COVID anxiety and compulsive exercise and ED pathology; COVID anxiety was more strongly related to compulsive exercise and ED pathology for individuals with lower intolerance of uncertainty | 3 |
| Schlegl, Maier, et al. (2020) | Germany | 159 (100%) | 22.4 (8.7) | Not reported | Not reported | Not reported | AN | Cross‐sectional (retrospective) | Psychological consequences of COVID | >70% reported that eating, shape and weight concerns, drive for physical activity, loneliness, sadness, and inner restlessness increased from pre‐ to during COVID and access to in person care decreased; participants reported daily routines, day planning, and enjoyable activities as the most helpful coping strategies; reduction in overall ED symptoms/taking on responsibility to recover, reduction in specific ED symptoms, more flexibility regarding meals and foods, wake‐up call/will to live, trying out therapy content, and accepting uncertainty in life were reported as positive impacts of COVID | 3 |
| Schlegl, Meule, et al. (2020) | Germany | 55 (100%) | 24.4 (6.4) | Not reported | Not reported | Not reported | BN | Cross‐sectional (retrospective) | Psychological consequences of COVID | 49% reported deterioration of ED symptomatology and 62% reported reduced quality of life; frequency of binge eating increased in 47% of patients, self‐induced vomiting in 36%, laxative use in 9%, and diuretic abuse in 7%; face‐to‐face psychotherapy decreased by 56%, videoconferencing therapy was used by 22% of patients; enjoyable activities, virtual contact with friends, and mild physical activity rated as most helpful coping strategies | 3 |
| Serin and Koç (2020) | Turkey | 1064 (59%) | Not reported | Not reported | Not reported | Not reported | n/a | Cross‐sectional | Eating behaviors; depression | External eating, but not emotional or restrictive eating, higher in participants who reported self‐isolating, compared to those who did not (>women) | 2 |
| Spettigue et al. (2021) | Canada | 48 (gender: 83% ciswomen; 2% transwomen; 4% transmen) | 14.6 (1.8) | Not reported | Not reported | Not reported | AN‐R n = 24; AN‐BP n = 7; ARFID n = 7; AAN n = 6; BN n = 1; UFED n = 3 | Cohort | ED examination; eating attitudes; clinical impairment | 40% cited pandemic as trigger for ED; inpatient admissions, emergency room consultation requests, and outpatient referrals deemed “urgent” were higher during COVID compared to pre‐COVID; compared to 2019 ED patients, ED patients in 2020 reported worse clinical impairment from ED symptoms (d = 0.44) and higher levels of eating restraint (d = 0.63) | 2 |
| Stoddard (2021) b | United States | 69 (gender identity: 96% women; 4% nonbinary) | 96% 18–34 years | 91% White; 7% Black; 3% Asian; 6% Latina; 3% “other” | Not reported | Not reported | Previous/current: 83/71% BI disturbance; 84/51% DE habits; 72/13% ED | Cross‐sectional | ED/BI status/concern level; impact of COVID; experiences of ED/BI; coping mechanisms | 84% reported concern about how their weight/bodies might be affected by COVID; 59% reported that their BI/ED worsened from pre‐ to post‐COVID; 78% reported that their relationships with their bodies have changed (70% negative, 26% positive, 4% neutral) | 2 |
| Swami, Horne, et al. (2021) | United Kingdom | 506 (gender identity: 44% women) | 34.3 (11.4) | 88.5% White | 89.1% heterosexual | Education: 10.9% high school; 27.9% advanced qualification; 38.3% undergraduate degree; 19.0% postgraduate degree; 3.9% other | n/a | Cross‐sectional | Perceived stress; ED inventory; body attitude; COVID‐related stress and anxiety | In women, COVID‐related anxiety associated with body dissatisfaction and COVID‐related anxiety and stress was associated with drive for thinness; in men, COVID‐related anxiety was associated with low body fat dissatisfaction and COVID‐related anxiety and stress was associated with muscularity dissatisfaction | 2 |
| Swami, Todd, et al. (2021) | United Kingdom | 600 (gender identity: 49% women) | 34.6 (12.3) | 85% White; 9% Asian; 2% Black; 4% mixed race; 0.3% “other” | 87% heterosexual; 3% gay/lesbian; 8% bisexual; 2% identified with another orientation | Not reported | n/a | Cross‐sectional | BI disturbance; COVID‐related stress; self‐compassion | COVID‐related stress associated with greater BI disturbance, mediated by lower self‐compassion; self‐compassion did not moderate effects of stress on BI disturbance | 2 |
| Tabler et al. (2021) b | United States | 411 (74% women; 5% transgender, genderqueer, or nonbinary) | 28.5 (11.4) | 86% White; 14% Latinx | 71% heterosexual; 29% lesbian, gay, bisexual, or queer | 44% working class; 44% middle class; 12% upper middle class (self‐report) | n/a | Cross‐sectional | ED examination; pandemic‐related stress | Pandemic‐related stress associated with ED symptoms and perceived weight gain (>LGBTQ+ individuals) | 2 |
| Taquet et al. (2021) | United States | 5,186,451 (55%) | 15.4 (9.0) | Not reported | Not reported | Not reported | Patients with any ED diagnosis | Cross‐sectional (retrospective) | Incidence of ED diagnosis | Diagnostic incidence of EDs 15.3% higher in 2020 compared with previous 3 years; increase occurred solely in women, and primarily related to teenagers and AN; higher proportion of patients with EDs in 2020 had suicidal ideation or attempted suicide | 1 |
| Termorshuizen et al. (2020) | United States; Netherlands | United States/Netherland: 511/510 (gender identity: 95/98% women; 2/0.6% nonbinary/genderfluid/“other”) | United States: 30.6 (9.4); Netherlands: 90% 16–39 years | Not reported | Not reported | Not reported | AN n = 665; BN n = 295; BED n = 216; AAN n = 203; OSFED n = 192; purging disorder n = 47; ARFID n = 36; night‐eating syndrome n = 25 | Cross‐sectional | Impact of COVID on EDs, general physical and mental well‐being, and ED treatment | Participants with AN reported increased restriction and fears about being able to find foods consistent with meal plan from pre‐ to during COVID; participants with BN and BED reported increases in binge eating and urges to binge; participants reported positive effects of COVID including greater connection with family, more time for self‐care, and motivation to recover | 2 |
| Trott et al. (2021) | United Kingdom | 319 (84%) | 36.7 (11.8) | Not reported | Not reported | Not reported | n/a | Longitudinal | Body dysmorphic symptoms; eating attitudes; exercise addiction | ED symptomatology and exercise increased while exercise addiction decreased from pre to postlockdown; no changes in body dysmorphic symptoms from pre‐ to postlockdown | 2 |
| Vall‐Roqué et al. (2021) | Spain | 2601 (gender: 100% women) | 24.1 (5.0) | Not reported | Not reported | Not reported | n/a | Cross‐sectional (retrospective) | ED inventory; social network sites use; self‐esteem | Social media network site use increased from pre‐ to during lockdown and was associated with lower self‐esteem (g = 0.15 for 14–24 year olds), higher body dissatisfaction (g = −0.14 for 14–24 year olds), and higher drive for thinness (g = −0.18 for 14–24 year olds, g = −0.22 for 25–35 year olds) (>younger age) | 2 |
| Vitagliano et al. (2021) | United States | 89 (sex assigned at birth: 89% female) | 18.9 (2.9) | 78% White, non‐Hispanic; 8% Asian; 7% Multiracial; 4% Hispanic; 1% Black; 2% “other” | Not reported | Not reported | 84% restrictive ED diagnosis; 16% other ED diagnosis | Cross‐sectional | ED related concerns and motivation to recover; triggering environment; ED diagnosis | 63% reported concern for worsening of their ED due to a “triggering environment”; 74% reported an increase in ED thoughts, 77% reported anxiety, 73% reported depression, and 80% reported isolation they perceived to be related to COVID; 29% reported decrease in motivation to recover they perceived to be related to COVID; participants who reported concern for worsening of their ED due to a triggering environment expressed decreased motivation to recover (OR = 18.1) and increased ED thoughts (OR = 23.8) compared to those who did not report concern for worsening of their ED due to a triggering environment | 3 |
| Vuillier et al. (2021) b | United Kingdom | 207 (63%) | 30.0 (9.7) | 93.7% White British/Irish/Scottish/European; 5.3% Asian; 0.5% Black; 0.5% Arab | Not reported | Not reported | AN n = 91; BN n = 46; BED n = 44; OSFED n = 26 | Cross‐sectional | Depression, anxiety, stress; ED examination | 83.1% reported worsening of ED symptomatology, 7.7% reported no change, 6.8% reported improvements in ED symptomatology; changes did not differ based on diagnosis; changes to routine and physical activity and emotion difficulties were most important factors in predicting change in ED symptoms | 3 |
| White III (2021) | United States | 311 (70%) | Not reported | 33.1% African American; 55.3% White; 11.6% “other” | Not reported | Not reported | n/a | Cross‐sectional (retrospective) | Physical activity; self‐esteem; BI | Self‐esteem and BI worsened from pre‐ to during COVID | 3 |
| Zhou and Wade (2021) | Australia | 100 (100%) | 19.9 (2.0) | 88% White; 6% Asian; 6% “other” | Not reported | Not reported | n/a | Cohort (part of randomized controlled trial) | ED examination; BI acceptance; self‐compassion | Weight concerns (d = 0.46), DE (fasting, binge eating, vomiting, and driven exercise; d = 0.55), and negative affect (d = 0.40) increased from pre‐ to during COVID, all associated with moderate effect sizes | 2 |
Note: Study quality was assessed using the EPHPP guidelines as follows: 1 = “strong,” 2 = “moderate,” 3 = “weak.” Effect size interpretation: Cohen's d/Hedge's g: 0.2 = small, 0.5 = medium, 0.8 = large; η 2/η 2 p: 0.01 = small, 0.06 = medium, 0.14 = large; r 2: 0.01 = small, 0.09 = medium, 0.25 = large; odds ratio (OR): 1.68 = small, 3.47 = medium, 6.71 = large c ; Pearson's r: 0.1 = small, 0.3 = medium, 0.5 = large.
Abbreviations: AN, anorexia nervosa; AAN, atypical anorexia; AN‐BP, binging/purging anorexia subtype; AN‐R, restrictive anorexia subtype; ARFID, avoidant restrictive food intake disorder; BED, binge eating disorder; BI, body image; BN, bulimia nervosa; DE, disordered eating; ED, eating disorder; OSFED, other specified feeding or eating disorder; UFED, unspecified feeding or eating disorder.
a
The majority of included studies did not specify whether they assessed sex assigned at birth, gender, or gender identity; where this information is available, sex assigned at birth, gender, and gender identity are reported separately.
b
Signifies a multimethods paper. Information detailed here concerns the quantitative methods, analysis, and findings. See Table 2 for the qualitative characteristics of this study.
c
Chen, H., Cohen, P., & Chen, S. (2010). How big is a big odds ratio? Interpreting the magnitudes of odds ratios in epidemiological studies. Communications in Statistics—Simulation and Computation, 39(4), 860–864.