Key Points
Question
Is course delivery model (entirely online vs mix of online and in-person classes) associated with college students’ mental health?
Findings
In this cross-sectional study of a nationwide data set that included 59 250 full-time undergraduate students, those attending fully online classes reported higher levels of psychological distress than students attending a mix of online and in-person classes.
Meaning
The findings of this study suggest that educational institutions and policy makers should weigh the risks and benefits when making determinations regarding school setting and transitions to online classes.
Abstract
Importance
College students in the US have been heavily affected by the COVID-19 pandemic. In addition to increased rates of depression and anxiety, college students have faced unprecedented stressors, such as geographic relocation and abrupt conversion from in-person classes to online classes.
Objective
To study the association between course delivery model and psychological distress among US college students.
Design, Setting, and Participants
This cross-sectional analysis used national data from the American College Health Association–National College Health Assessment III data set. Data were gathered from a web-based survey administered from January to early June 2021 to full-time US college students attending 4-year programs.
Exposure
Course delivery model was self-reported.
Main Outcomes and Measures
Psychological distress was measured using the Kessler Screening Scale for Psychological Distress.
Results
This study evaluated 59 250 full-time undergraduate students (68.1% women; 51.5% White students; mean [SD] age, 21.2 [4.3] years); 3.5% attended fully in-person classes, 61.2% attended fully online classes, and 35.3% attended a mixed format of in-person and online classes. Students who attended classes fully online reported higher levels of psychological distress than those who attended a mix of online and in-person classes (b = 0.76 [99% CI, 0.64-0.88]; P < .001). This association remained significant after controlling for geographic region, year in school, gender, race and ethnicity, food security, current anxiety and/or depressive disorders, COVID-19 concerns, and residence (living on campus, off campus with family, or other off-campus arrangements) (b = 0.18 [99% CI, 0.04-0.31]; P = .001), as well as time spent socializing with friends (b = 0.13 [99% CI, 0.002-0.26]; P = .009).
Conclusions and Relevance
The findings of this study suggest that mental health professionals may wish to consider the association of course delivery models with mental health outcomes when working with college students. Colleges should be aware of the mental health burden associated with attending fully online classes and consider possible in-person components and supports for students.
This cross-sectional analysis uses national data from the American College Health Association–National College Health Assessment III data set to study the association between course delivery model and psychological distress among college students.
Introduction
College students worldwide are facing unprecedented stressors brought about by the COVID-19 pandemic.1,2 Many college students have endured the loss of a loved one, faced financial hardships, or experienced racial discrimination during the pandemic.3 Rates of depression and anxiety among US college students have increased markedly, with 6 of every 10 college students reporting symptoms of anxiety or depression during the pandemic.2,4 Compared with before the pandemic, the prevalence of depressive symptoms among US adults aged 18 to 39 years old during the pandemic more than quadrupled by April 2020.5 In addition, most US college students had to relocate from their college campuses1 within weeks from the declaration of the pandemic.6 A Pew Research Center analysis indicated that between February and July 2020, 2.1 million young adults between 18 and 24 years of age moved back in with their parents.7
Concomitant shifts in the learning environment during the pandemic, such as the transition to virtual classes, also altered course delivery models and structures.8 During the fall of 2020, 43% of 4-year colleges had fully online classes, 34% included a mix of in-person and online classes, and 13% had fully in-person classes.9 Challenges faced by college students in remote learning environments include limited internet or technology access, with negative consequences on academic performance.10 Challenges also include the loss of student experiences, such as extracurricular activities, internships, trips to study abroad, service learning, and social events.10 The deprivation of these milestone events, as well as the loss of normalcy, friendships, and connection with others, may contribute to personal distress.11 In addition, those taking courses online may include students living at home during the first year of the pandemic. Students’ residence—whether with peers or family—may predispose them to different socialization experiences or levels of distress. There is ample evidence that socializing with others is critical for supporting both mental and physical health.12,13 However, many mental health professionals may fail to consider the association of such social determinants with mental health outcomes in their clinical approach.14
The association between fully online classes and psychological distress—the set of cognitive, emotional, and behavioral symptoms associated with mental health disorders15—remains understudied among college students, to our knowledge. One small cross-sectional survey including fewer than 200 participants showed that most college students had difficulties adjusting to online learning and focusing on academic work during the pandemic and that academic challenges were associated with higher rates of depression and anxiety.16 Of the studies that have been conducted on the association of online learning with student outcomes, most focused on academic outcomes.17,18 To our knowledge, there are no large-scale studies before or during the COVID-19 pandemic that have examined the association between course delivery model (entirely online vs mix of online and in-person classes) and college students’ mental health.
To address this gap, our study analyzed a nationwide sample of undergraduate students in the US from spring 2021 to measure the prevalence of college students who were engaged in course delivery models that were online only, in-person only, and mixed (online and in-person). We examined whether students attending online classes reported higher rates of psychological distress compared with students attending mixed online and in-person classes.
Methods
Data Source and Sample
This cross-sectional study was based on the American College Health Association–National College Health Assessment III (ACHA-NCHA),19 a biannual survey administered to students in higher educational institutions across the US. The ACHA-NCHA requires institutions either to have all students respond or to randomly select a sample of students. The spring 2021 survey, administered from January to early June 2021, was entirely web-based and included demographic data, psychometric scales, and COVID-19–related questions. Our analysis was based on 59 250 full-time undergraduate students attending 4-year US colleges or universities during spring 2021 with data available on all measures as described. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for reporting observational studies.20 The use of this existing and deidentified data set from the ACHA was approved as an expedited application through the institutional review board at Mass General Brigham. The institutional review board of Mass General Brigham deemed this analysis exempt from human participants review as it used deidentified data from the ACHA-NCHA. Participant consent was first implied when participants clicked on the link within an email message to access the survey, with procedures approved by the institutional review board of the students’ institution. Second, participants were presented with information and instructions on the first page of ACHA-NCHA, including that by clicking “Begin Survey,” they consented to participate in the survey.
Measures
Exposures: Course Delivery Model and Place of Residence
The course delivery model variable was assessed using the following question: “I am taking classes this term,” with response options of “entirely in-person,” “entirely online,” or “a mix of in-person and online classes.” The place of residence was assessed using the following question: “Where do you currently live?” Students who answered “in a fraternity or sorority residence” or “campus or university housing” were coded as the on-campus group. Those who answered “parent/guardian/other family member’s home” were coded as the off-campus with family group. Students who answered “off-campus or other non-university housing” were coded as the other off-campus arrangements group.
Outcome: Psychological Distress
Psychological distress was measured using the Kessler Screening Scale for Psychological Distress.15 The scale consists of 6 questions, each of which starts with, “During the past 30 days, about how often did you feel…” The questions asked about one’s experience of being “nervous,” “hopeless,” “restless or fidgety,” “so sad nothing could cheer you up,” “that everything was an effort,” and “worthless.” Each question was answered on a 5-point Likert scale, where 0 indicated “none of the time” and 4 indicated “all of the time,” with a total score range of 0 to 24. Higher scores correspond to greater psychological distress. In line with previous studies,21,22 psychological distress was examined as a continuous variable in our analysis. The Cronbach α for these items in our sample was 0.89, indicating good reliability.
Covariates
A detailed description of our covariates (sociodemographic characteristics, current anxiety and/or depressive disorders, socializing time, and COVID-19 concerns) can be found in the eAppendix in the Supplement.
Statistical Analysis
We used Stata, version 17 (StataCorp LLC) for our data analysis.23 In line with previous research,3,24 data cleaning was performed by removing observations with scores outside the plausible range for height (<47.2 inches [120 cm] or >94 inches [239 cm]), weight (<75 lb [34 kg] or >600 lb [272 kg]), and body mass index (>80; calculated as weight in kilograms divided by height in meters squared). Participants who reported not currently having a place to live or temporarily staying with a friend were removed from the analysis (0.1% excluded). Participants with missing data on any of the variables used in the final model were also excluded (5.2% excluded). The resulting sample size was 59 250 full-time undergraduate students in 4-year US colleges or universities. This sample was used for our descriptive analysis in Table 1.
Table 1. General Demographic and Key Characteristics of the Sample.
| Characteristic | Students, No. (%)a | ||||||
|---|---|---|---|---|---|---|---|
| Total (N = 59 250 [100%]) | Course delivery modelb | Residenceb | |||||
| In person (n = 2075 [3.5%]) | Online (n = 36 273 [61.2%]) | Mixed (n = 20 902 [35.3%]) | On campus (n = 16 887 [28.5%]) | Off campus with family (n = 22 074 [37.3%]) | Other off-campus arrangements (n = 20 289 [34.2%]) | ||
| Region | |||||||
| Northeast | 6932 (11.7) | 293 (14.1) | 4070 (11.2) | 2569 (12.3) | 2809 (16.6) | 2404 (10.9) | 1719 (8.5) |
| Midwest | 13 684 (23.1) | 863 (41.6) | 4431 (12.2) | 8390 (40.1) | 6940 (41.1) | 1994 (9.0) | 4750 (23.4) |
| South | 5304 (9.0) | 472 (22.7) | 1494 (4.1) | 3338 (16.0) | 2203 (13.0) | 864 (3.9) | 2237 (11.0) |
| West | 33 330 (56.3) | 447 (21.5) | 26 278 (72.4) | 6605 (31.6) | 4935 (29.2) | 16 812 (76.2) | 11 583 (57.1) |
| Age, mean (SD), y | 21.2 (4.3) | 20.6 (2.9) | 21.6 (4.8) | 20.6 (3.3) | 19.7 (2.0) | 20.7 (3.0) | 23.0 (5.9) |
| Current anxiety disorder | 11 690 (19.7) | 368 (17.7) | 6623 (18.3) | 4699 (22.5) | 3618 (21.4) | 3275 (14.8) | 4797 (23.6) |
| Current depressive disorder | 9445 (15.9) | 264 (12.7) | 5441 (15.0) | 3740 (17.9) | 2874 (17.0) | 2613 (11.8) | 3958 (19.5) |
| COVID-19 concerns, mean (SD)c | 14.0 (5.8) | 11.7 (5.8) | 14.7 (5.7) | 12.9 (5.7) | 13.1 (5.6) | 14.8 (5.7) | 13.9 (5.8) |
| Race and ethnicity | |||||||
| American Indian | 260 (0.4) | 16 (0.8) | 139 (0.4) | 105 (0.5) | 75 (0.4) | 74 (0.3) | 111 (0.5) |
| Asian | 8804 (14.9) | 130 (6.3) | 7030 (19.4) | 1644 (7.9) | 1634 (9.7) | 4976 (22.5) | 2194 (10.8) |
| Black | 1857 (3.1) | 46 (2.2) | 1198 (3.3) | 613 (2.9) | 665 (3.9) | 667 (3.0) | 525 (2.6) |
| Hispanic | 9061 (15.3) | 78 (3.8) | 7767 (21.4) | 1216 (5.8) | 1181 (7.0) | 5764 (26.1) | 2116 (10.4) |
| Middle Eastern | 598 (1.0) | 6 (0.3) | 494 (1.4) | 98 (0.5) | 85 (0.5) | 361 (1.6) | 152 (0.7) |
| Multiracial | 7634 (12.9) | 182 (8.8) | 5218 (14.4) | 2234 (10.7) | 2064 (12.2) | 2976 (13.5) | 2594 (12.8) |
| Native Hawaiian | 145 (0.2) | 2 (0.1) | 118 (0.3) | 25 (0.1) | 20 (0.1) | 89 (0.4) | 36 (0.2) |
| White | 30 490 (51.5) | 1602 (77.2) | 14 042 (38.7) | 14 846 (71.0) | 11 062 (65.5) | 7003 (31.7) | 12 425 (61.2) |
| Otherd | 401 (0.7) | 13 (0.6) | 267 (0.7) | 121 (0.6) | 101 (0.6) | 164 (0.7) | 136 (0.7) |
| Gender | |||||||
| Men | 16 642 (28.1) | 694 (33.4) | 9588 (26.4) | 6360 (30.4) | 5138 (30.4) | 5859 (26.5) | 5645 (27.8) |
| Women | 40 327 (68.1) | 1334 (64.3) | 25 261 (69.6) | 13 732 (65.7) | 10 980 (65.0) | 15 448 (70.0) | 13 899 (68.5) |
| Other | 2281 (3.8) | 47 (2.3) | 1424 (3.9) | 810 (3.9) | 769 (4.6) | 767 (3.5) | 745 (3.7) |
| International student | 3522 (6.1) | 87 (4.3) | 2422 (6.9) | 1013 (5.0) | 969 (6.0) | 1465 (6.8) | 1088 (5.5) |
| Year in school | |||||||
| First | 14 753 (24.9) | 540 (26.0) | 8332 (23.0) | 5881 (28.1) | 7452 (44.1) | 5983 (27.1) | 1318 (6.5) |
| Second | 12 822 (21.6) | 466 (22.5) | 7239 (20.0) | 5117 (24.5) | 4400 (26.1) | 4687 (21.2) | 3735 (18.4) |
| Third | 15 984 (27.0) | 457 (22.0) | 10 365 (28.6) | 5162 (24.7) | 3016 (17.9) | 5922 (26.8) | 7046 (34.7) |
| Fourth | 12 267 (20.7) | 541 (26.1) | 7818 (21.6) | 3908 (18.7) | 1830 (10.8) | 3979 (18.0) | 6458 (31.8) |
| Fifth or more | 3424 (5.8) | 71 (3.4) | 2519 (6.9) | 834 (4.0) | 189 (1.1) | 1503 (6.8) | 1732 (8.5) |
| Food security | |||||||
| High | 37 973 (64.1) | 1385 (66.7) | 23 508 (64.8) | 13 080 (62.6) | 10 641 (63.0) | 15 589 (70.6) | 11 743 (57.9) |
| Low | 14 953 (25.2) | 509 (24.5) | 8832 (24.3) | 5612 (26.8) | 4507 (26.7) | 4800 (21.7) | 5646 (27.8) |
| Very low | 6324 (10.7) | 181 (8.7) | 3933 (10.8) | 2210 (10.6) | 1739 (10.3) | 1685 (7.6) | 2900 (14.3) |
| Socializing time (h/wk) | |||||||
| Low (0) | 5657 (9.5) | 70 (3.4) | 4459 (12.3) | 1128 (5.4) | 771 (4.6) | 3136 (14.2) | 1750 (8.6) |
| Medium (1-5) | 24 350 (41.1) | 677 (32.6) | 16 268 (44.8) | 7405 (35.4) | 5309 (31.4) | 10 916 (49.5) | 8125 (40.0) |
| High (≥6) | 29 243 (49.4) | 1328 (64.0) | 15 546 (42.9) | 12 369 (59.2) | 10 807 (64.0) | 8022 (36.3) | 10 414 (51.3) |
| Place of residence | |||||||
| On campus | 16 887 (28.5) | 1287 (62.0) | 5101 (14.1) | 10 499 (50.2) | NA | NA | NA |
| Off campus with family | 22 074 (37.3) | 146 (7.0) | 19 290 (53.2) | 2638 (12.6) | NA | NA | NA |
| Other off-campus arrangements | 20 289 (34.2) | 642 (30.9) | 11 882 (32.8) | 7765 (37.1) | NA | NA | NA |
| Course delivery model | |||||||
| In-person | 2075 (3.5) | NA | NA | NA | 1287 (7.6) | 146 (0.7) | 642 (3.2) |
| Online | 36 273 (61.2) | NA | NA | NA | 5101 (30.2) | 19 290 (87.4) | 11 882 (58.6) |
| Mixed | 20 902 (35.3) | NA | NA | NA | 10 499 (62.2) | 2638 (12.0) | 7765 (38.3) |
Abbreviation: NA, not applicable.
Percentages may not total 100% because of rounding.
The χ2 test and analysis of variance examining characteristics by course delivery model and by residence showed an overall statistically significant difference in distribution (P < .001).
COVID-19 concerns scale is composed of 6 items. Each item was measured on a 5-point Likert scale, where 0 indicated “not concerned at all” and 4 indicated “extremely concerned,” with a total score range from 0 to 24.
Students who indicated that their racial or ethnic identity is not listed were recoded as “Other.”
Students attending entirely in-person classes were excluded from the regression analyses given their small sample size (n = 2075 [3.5%]). The sample size for our regression analyses was 57 175 students, comprising those who attended fully online classes and those who attended mixed online and in-person classes. The first linear regression model examined psychological distress as an outcome based on course delivery model as an independent variable (block 1). The second model adjusted for covariates through a multiple linear regression analysis (block 2). The third model included the socialization variable (block 3) (Table 1). A simple linear regression model was used for Table 2 block 1. Multiple linear regression models were used for Table 2 blocks 2 and 3. All statistical tests were 2-sided. Consistent with prior published research using ACHA-NCHA data,25 a significance level of P < .01 and 99% CIs were used given the large sample size and number of comparisons being made.
Table 2. Linear Regression Model With Psychological Distress as an Outcome and Course Delivery Model as an Independent Variable (Online Only vs Mixed Format)a.
| Variable | Block 2b | Block 3c | ||||
|---|---|---|---|---|---|---|
| Unstandardized coefficient (99% CI) | Standardized β coefficient | P value | Unstandardized coefficient (99% CI) | Standardized β coefficient | P value | |
| Region | ||||||
| Northeast | 0 [Reference] | 0 | NA | 0 [Reference] | 0 | NA |
| Midwest | −0.03 (−0.22 to 0.17) | −0.002 | .72 | 0.002 (−0.19 to 0.19) | 0.000 | .98 |
| South | −0.20 (−0.44 to 0.04) | −0.01 | .04 | −0.164 (−0.403 to 0.075) | −0.008 | .08 |
| West | 0.114 (−0.06 to 0.29) | 0.01 | .09 | 0.07 (−0.10 to 0.24) | 0.006 | .30 |
| Year in school | ||||||
| First | 0 [Reference] | 0 | NA | 0 [Reference] | 0 | NA |
| Second | 0.06 (−0.10 to 0.21) | 0.004 | .33 | 0.08 (−0.08 to 0.24) | 0.006 | .18 |
| Third | −0.27 (−0.43 to −0.12) | −0.02 | <.001 | −0.28 (−0.43 to −0.13) | −0.02 | <.001 |
| Fourth | −0.60 (−0.76 to −0.43) | −0.04 | <.001 | −0.60 (−0.76 to −0.43) | −0.04 | <.001 |
| Fifth or more | −0.56 (−0.81 to −0.31) | −0.02 | <.001 | −0.65 (−0.89 to −0.40) | −0.03 | <.001 |
| Gender | ||||||
| Men | 0 [Reference] | 0 | NA | 0 [Reference] | 0 | NA |
| Women | 0.47 (0.35 to 0.589) | 0.04 | <.001 | 0.45 (0.33 to 0.57) | 0.04 | <.001 |
| Other | 2.60 (2.31 to 2.88) | 0.09 | <.001 | 2.57 (2.28 to 2.85) | 0.09 | <.001 |
| Race and ethnicity | ||||||
| White | 0 [Reference] | 0 | NA | 0 [Reference] | 0 | NA |
| American Indian | −0.002 (−0.80 to 0.80) | 0.000 | .995 | −0.18 (−0.98 to 0.62) | −0.002 | .56 |
| Asian | 0.70 (0.54 to 0.86) | 0.05 | <.001 | 0.68 (0.52 to 0.84) | 0.04 | <.001 |
| Black | 0.04 (−0.27 to 0.34) | 0.001 | .76 | −0.11 (−0.42 to 0.19) | −0.004 | .33 |
| Hispanic | −0.11 (−0.28 to 0.05) | −0.007 | .08 | −0.23 (−0.40 to −0.07) | −0.02 | <.001 |
| Middle Eastern | 1.35 (0.83 to 1.87) | 0.03 | <.001 | 1.30 (0.78 to 1.82) | 0.02 | <.001 |
| Multiracial | 0.39 (0.22 to 0.55) | 0.02 | <.001 | 0.35 (0.18 to 0.51) | 0.02 | <.001 |
| Native Hawaiian | 0.15 (−0.90 to 1.19) | 0.001 | .72 | 0.10 (−0.95 to 1.14) | 0.001 | .81 |
| Otherd | 0.32 (−0.32 to 0.96) | 0.005 | .20 | 0.15 (−0.48 to 0.79) | 0.002 | .54 |
| Food security | ||||||
| High | 0 [Reference] | 0 | NA | 0 [Reference] | 0 | NA |
| Low | 1.52 (1.39 to 1.64) | 0.12 | <.001 | 1.51 (1.39 to 1.64) | 0.12 | <.001 |
| Very low | 3.16 (2.99 to 3.34) | 0.18 | <.001 | 3.12 (2.94 to 3.29) | 0.18 | <.001 |
| Current anxiety disorder | 1.03 (0.84 to 1.22) | 0.08 | <.001 | 1.05 (0.86 to 1.24) | 0.08 | <.001 |
| Current depressive disorder | 2.76 (2.56 to 2.97) | 0.18 | <.001 | 2.72 (2.51 to 2.93) | 0.18 | <.001 |
| COVID-19 concerns | 0.25 (0.24 to 0.26) | 0.26 | <.001 | 0.25 (0.24 to 0.25) | 0.26 | <.001 |
| Place of residence | ||||||
| On campus | 0 [Reference] | 0 | NA | 0 [Reference] | 0 | NA |
| Off campus with family | 0.30 (0.14 to 0.45) | 0.03 | <.001 | 0.12 (−0.03 to 0.28) | 0.01 | .04 |
| Other off-campus arrangements | −0.54 (−0.69 to −0.39) | −0.05 | <.001 | −0.61 (−0.76 to −0.46) | −0.05 | <.001 |
| Course delivery model | ||||||
| Mixed | 0 [Reference] | 0 | NA | 0 [Reference] | 0 | NA |
| Online | 0.18 (0.04 to 0.31) | 0.02 | .001 | 0.13 (0.002 to 0.26) | 0.01 | .009 |
| Socializing time | ||||||
| Low | NA | NA | NA | 0 [Reference] | 0 | NA |
| Medium | NA | NA | NA | −1.18 (−1.37 to −0.99) | −0.11 | <.001 |
| High | NA | NA | NA | −1.73 (−1.92 to −1.55) | −0.16 | <.001 |
| R 2 | 0.229 | 0.237 | ||||
Abbreviation: NA, not applicable.
Block 1 was unadjusted (course delivery model variable only). The unstandardized coefficient for online-only relative to mixed format was 0.76 (99% CI, 0.64-0.88), with a standardized β coefficient of 0.07 (P < .001) and R2 = 0.004.
Adjusted (course delivery model and all single variables except socializing time).
Adjusted (course delivery model, all single variables, and socializing time).
Students who indicated that their racial or ethnic identity is not listed were recoded as “Other.”
Results
This study evaluated 59 250 full-time undergraduate students (68.1% women; 51.5% White students; mean [SD] age, 21.2 [4.3] years). Table 1 summarizes the general demographic and key characteristics of the full sample. More than half the participants (64.1%) reported a high level of food security. Almost one-fifth of the participants (19.7%) reported having a current anxiety disorder, and 15.9% reported having a current depressive disorder.
Of the 59 250 participants in our sample, 3.5% attended fully in-person classes, 61.2% attended fully online classes, and 35.3% attended a mixed format of in-person and online classes. Of the 59 250 participants, 28.5% lived on campus, 37.3% lived off campus with family, and 34.2% lived in other off-campus arrangements. About half the college students (49.4%) reported socializing 6 or more hours a week, while 41.1% of students spent 1 to 5 hours a week socializing. A total of 9.5% of students reported spending 0 hours a week socializing with friends. These rates appear to vary widely based on the various course delivery models. Students attending online-only classes had the lowest socialization levels, with 12.3% of students reporting no socializing compared with 5.4% of students attending mixed format classes and 3.4% of students attending fully in-person classes.
Table 2 presents results of the linear regression models with psychological distress as an outcome and course delivery model as an independent variable (online-only vs mixed-format classes). Compared with the students attending mixed-format classes, those who attended fully online classes reported greater distress (b = 0.76 [99% CI, 0.64-0.88]; P < .001; block 1). This association remained significant after controlling for region, year in school, gender, race and ethnicity, food security, current anxiety and/or depressive disorders, COVID-19 concerns, and place of residence (b = 0.18 [99% CI, 0.04-0.31]; P = .001; block 2). Even when controlling for socializing time, the association between attending classes online and increased distress levels remained significant (b = 0.13 [99% CI, 0.002-0.26]; P = .009; block 3).
Students who lived in other off-campus arrangements reported less distress (b = −0.54 [99% CI, −0.69 to −0.39]; P < .001; block 2) relative to those who lived on-campus; this association remained significant after controlling for socializing time (b = −0.61 [99% CI, −0.76 to −0.46]; P < .001; block 3). Students who lived off campus with family reported more distress (b = 0.30 [99% CI, 0.14-0.45]; P < .001; block 2) relative to those who lived on campus, but this association did not remain significant after controlling for socializing time (b = 0.12 [99% CI, −0.03 to 0.28]; P = .04; block 3) (Table 2).
Discussion
Most (61.2%) of the 59 250 US college students in our sample attended classes fully online, 35.3% attended a mixed format of in-person and online classes, and 3.5% attended fully in-person classes. Based on our analyses, it appears that students whose classes were offered entirely online were at risk for increased psychological distress compared with those attending a mix of in-person and online classes. This association remained significant even after controlling for geographic region and a wide range of sociodemographic characteristics, as well as when controlling for students’ reported amount of time spent socializing with friends.
Why might a course delivery model with some in-person experiences be more beneficial than a fully online course delivery model? First, the culture of student life is significantly altered with an online-only format, whereas some degree of normalcy may be preserved with a mixed format that includes at least some in-person experiences. The COVID-19 pandemic’s negative association with school culture formation was a frequently reported concern in a qualitative study of 43 primary and secondary school educators.26 Socializing with friends was likely more challenging for those who attended classes only online, as such engagement requires greater intentionality and effort. In contrast, a mixed format still afforded at least some in-person experiences that students were accustomed to, with informal opportunities for social interaction. Relatedly, those who attended classes only online were likely to have altered, limited, or no opportunities for participating in extracurricular activities.27 Students who attended classes in a mixed format might also have had the choice of attending online or in-person, affording students the flexibility of attending in the format that is most convenient on any particular day.28 Such increased perceived control could also help mitigate the negative effect of stressful situations.29
Second, students who attended classes only online may have experienced greater distress from academic challenges.30,31,32 Factors other than socialization were likely at play given that the association between course delivery model and psychological distress holds even after controlling for reported socializing time. A 2022 mixed-methods study observed that college and graduate students found it challenging to engage during online classes.33 Some of the challenges reported included feeling distracted and procrastinating.33 Students may experience decreased motivation to engage with faculty when attending class fully online compared with when there is an opportunity for face-to-face interactions.34 Online classes may be held asynchronously, which might also increase the burden of time management, as students would be expected to go through course content on their own.34 Teaching methods were likely altered in the transition to online formats; changes in teaching methods were found to be a major source of academic-related frustration among college students during the pandemic.35
The shift to online classes was intended to limit in-person contact via social distancing,36,37 a key strategy for mitigating COVID-19 transmission.38 Despite the protections for physical health afforded by these strategies, our results suggest that fully online classes could be associated with worse mental health. Although a mixed format could present additional logistical complications for educators, our findings suggest that some amount of in-person instructional time may be protective for students’ mental health.
Although our primary study aim was to investigate the association between course delivery model and psychological distress, our secondary analyses revealed an association between students’ place of residence and psychological distress. Students who lived in other off-campus arrangements reported lower levels of psychological distress compared with those who lived on campus. After accounting for socializing time, students living off campus with family did not report different psychological distress levels than those who lived on campus. This finding addresses the current literature that shows mixed findings on the protective or harmful associations between living with family and the mental health of college students. Lee et al2 found that more than one-third of students had strained family relationships because of the COVID-19 pandemic, and most of these students found it harder to complete the semester at home. On the other hand, Davitt et al39 found that college and university students living with a parent or guardian during the pandemic had less food insecurity, less need to work, lower stress, improved health status, and more home-cooked meals compared with students living on their own. Students who moved residences because of the pandemic, many back to their parents’ homes, were also found to have a greater reduction in alcohol consumption than students who did not move.40 Our data indicate that socialization with peers may be an additional factor to consider when assessing how living with family is associated with the mental health of college students.
Limitations
The findings reported in this study have limitations, including several associated with the study design. First, the nature and extent of the in-person component of the mixed course delivery models are unknown and could vary by school. For example, mixed course delivery models can refer to programs with some online and some in-person components or to programs that have a hybrid model in which some students attend online while others attend in person. On an individual level, mixed course delivery models may also vary. For instance, students who reported having mixed course delivery models may have had one class in person and the rest online, or they may have experienced all classes as hybrid. In addition, some students may have had a choice in the course delivery at any given time. Second, the study’s cross-sectional design means that both the exposure and the outcomes were measured simultaneously, limiting causal inference. Third, the survey was administered online, and the measures were self-reported. Self-reported measures may be affected by recall bias and misinterpretation. Fourth, our study is limited by the available variables in the survey. For example, given the lack of data on socioeconomic status, food security was used as a proxy; however, it may not be an accurate measure of the socioeconomic status of college students. Similarly, while place of residence was included in our analysis, there may have been other factors about college students’ living arrangements that were not considered. In addition, there may be other potentially significant factors associated with college students’ well-being that we did not investigate. Future studies can examine how much of an in-person component is needed when a student engages in a mixed model to offset the mental health cost of online-only classes. In addition, it would be important to understand the extent to which these results hold under various pandemic conditions (ie, during a surge in cases) or under nonpandemic conditions.
Conclusions
Although the shift to online college classes has been shown to be feasible and arguably necessary in the context of the COVID-19 pandemic,6,41 our study suggests a potential negative association between such a shift and college students’ mental health. Our results have implications for educational institutions and policy makers weighing the risks and benefits when making determinations regarding school setting and transitions to online classes. Although online classes may be simpler logistically and may minimize the risk of COVID-19 transmission, they also may increase the risk of negative mental health sequelae that should not be ignored.
Our analysis also offers new insights regarding the association of widely scaled student educational experiences with individual psychological distress. A question that emerges is whether these same results would be maintained if online courses continued for a longer period of time (ie, would adaptations be made by students that would eventually mitigate psychological distress?) or if students had a choice in the way they take their courses. Finally, these results are particularly relevant to mental health professionals within educational settings. Knowing that a student is attending classes fully online may provide insight that informs therapeutic approaches and suggestions for recovery.
eAppendix. Covariates
eReferences.
References
- 1.Conrad RC, Hahm HC, Koire A, Pinder-Amaker S, Liu CH. College student mental health risks during the COVID-19 pandemic: implications of campus relocation. J Psychiatr Res. 2021;136:117-126. doi: 10.1016/j.jpsychires.2021.01.054 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lee J, Solomon M, Stead T, Kwon B, Ganti L. Impact of COVID-19 on the mental health of US college students. BMC Psychol. 2021;9(1):95. doi: 10.1186/s40359-021-00598-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.ElTohamy A, Hyun S, Macaranas AR, Chen JA, Stevens C, Liu CH. Testing positive, losing a loved one, and financial hardship: real-world impacts of COVID-19 on US college student distress. J Affect Disord. 2022;314:357-364. doi: 10.1016/j.jad.2022.07.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Liu CH, Zhang E, Wong GTF, Hyun S, Hahm HC. Factors associated with depression, anxiety, and PTSD symptomatology during the COVID-19 pandemic: clinical implications for U.S. young adult mental health. Psychiatry Res. 2020;290:113172. doi: 10.1016/j.psychres.2020.113172 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ettman CK, Abdalla SM, Cohen GH, Sampson L, Vivier PM, Galea S. Prevalence of depression symptoms in US adults before and during the COVID-19 pandemic. JAMA Netw Open. 2020;3(9):e2019686. doi: 10.1001/jamanetworkopen.2020.19686 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Camargo CP, Tempski PZ, Busnardo FF, Martins MA, Gemperli R. Online learning and COVID-19: a meta-synthesis analysis. Clinics (Sao Paulo). 2020;75:e2286. doi: 10.6061/clinics/2020/e2286 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Fry R, Passel JS, Cohn D. A majority of young adults in the U.S. live with their parents for the first time since the Great Depression. Pew Research Center. September 4, 2020. Accessed November 26, 2021. https://www.pewresearch.org/fact-tank/2020/09/04/a-majority-of-young-adults-in-the-u-s-live-with-their-parents-for-the-first-time-since-the-great-depression/
- 8.Cameron M, Lacy TA, Siegel P, et al. 2019–20 National Postsecondary Student Aid Study (NPSAS:20): first look at the impact of the coronavirus (COVID-19) pandemic on undergraduate student enrollment, housing, and finances (preliminary data). National Center for Education Statistics. June 16, 2021. Accessed March 7, 2022. https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2021456
- 9.National Center for Education Statistics . Table 3: number and percentage distribution of students enrolled at Title IV institutions, by control of institution, student level, level of institution, distance education status of student, and distance education status of institution: United States, fall 2020. Published 2021. Accessed March 7, 2022. https://nces.ed.gov/ipeds/search/ViewTable?tableId=29450
- 10.Lederer AM, Hoban MT, Lipson SK, Zhou S, Eisenberg D. More than inconvenienced: the unique needs of U.S. college students during the COVID-19 pandemic. Health Educ Behav. 2021;48(1):14-19. doi: 10.1177/1090198120969372 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sirrine EH, Kliner O, Gollery TJ. College student experiences of grief and loss amid the COVID-19 global pandemic. Omega (Westport). Published online June 23, 2021. doi: 10.1177/00302228211027461 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Martino J, Pegg J, Frates EP. The connection prescription: using the power of social interactions and the deep desire for connectedness to empower health and wellness. Am J Lifestyle Med. 2015;11(6):466-475. doi: 10.1177/1559827615608788 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Alegría M, NeMoyer A, Falgàs Bagué I, Wang Y, Alvarez K. Social determinants of mental health: where we are and where we need to go. Curr Psychiatry Rep. 2018;20(11):95. doi: 10.1007/s11920-018-0969-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Bhalla IP, Stefanovics EA, Rosenheck RA. Social determinants of mental health care systems: intensive community based care in the Veterans Health Administration. BMC Public Health. 2020;20(1):1311. doi: 10.1186/s12889-020-09402-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kessler RC, Barker PR, Colpe LJ, et al. Screening for serious mental illness in the general population. Arch Gen Psychiatry. 2003;60(2):184-189. doi: 10.1001/archpsyc.60.2.184 [DOI] [PubMed] [Google Scholar]
- 16.Kecojevic A, Basch CH, Sullivan M, Davi NK. The impact of the COVID-19 epidemic on mental health of undergraduate students in New Jersey, cross-sectional study. PLoS One. 2020;15(9):e0239696. doi: 10.1371/journal.pone.0239696 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Figlio D, Rush M, Yin L. Is it live or is it internet? experimental estimates of the effects of online instruction on student learning. J Labor Econ. 2013;31(4):763-784. doi: 10.1086/669930 [DOI] [Google Scholar]
- 18.Joyce T, Crockett S, Jaeger DA, Altindag O, O’Connell SD. Does classroom time matter? Econ Educ Rev. 2015;46:64-77. doi: 10.1016/j.econedurev.2015.02.007 [DOI] [Google Scholar]
- 19.American College Health Association . American College Health Association–National College Health Assessment III: Undergraduate Student Reference Group Data Report. American College Health Association; 2021. [Google Scholar]
- 20.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453-1457. doi: 10.1016/S0140-6736(07)61602-X [DOI] [PubMed] [Google Scholar]
- 21.Keum BTH, Cano MÁ. Online racism, psychological distress, and alcohol use among racial minority women and men: a multi-group mediation analysis. Am J Orthopsychiatry. 2021;91(4):524-530. doi: 10.1037/ort0000553 [DOI] [PubMed] [Google Scholar]
- 22.Mitchell CM, Beals J. The utility of the Kessler Screening Scale for Psychological Distress (K6) in two American Indian communities. Psychol Assess. 2011;23(3):752-761. doi: 10.1037/a0023288 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.StataCorp. Stata statistical software. Release 17. StataCorp LLC; 2021. Accessed November 13, 2021. https://www.stata.com/
- 24.Littman AJ, Boyko EJ, McDonell MB, Fihn SD. Evaluation of a weight management program for veterans. Prev Chronic Dis. 2012;9:E99. doi: 10.5888/pcd9.110267 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Liu CH, Stevens C, Wong SHM, Yasui M, Chen JA. The prevalence and predictors of mental health diagnoses and suicide among U.S. college students: implications for addressing disparities in service use. Depress Anxiety. 2019;36(1):8-17. doi: 10.1002/da.22830 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Koç S, Koç A. The effect failing to perform extracurricular activities has had on school culture and values education during the COVID-19 pandemic. Front Psychol. 2021;12:778678. doi: 10.3389/fpsyg.2021.778678 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Idris F, Zulkipli IN, Abdul-Mumin KH, et al. Academic experiences, physical and mental health impact of COVID-19 pandemic on students and lecturers in health care education. BMC Med Educ. 2021;21(1):542. doi: 10.1186/s12909-021-02968-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Drea J. Online? in person? the power of letting students choose. Harvard Business Publishing. Published March 19, 2021. Accessed March 31, 2022. https://hbsp.harvard.edu/inspiring-minds/online-in-person-the-power-of-letting-students-choose
- 29.Paterson RJ, Neufeld RWJ. What are my options?: influences of choice availability on stress and the perception of control. J Res Pers. 1995;29(2):145-167. doi: 10.1006/jrpe.1995.1009 [DOI] [Google Scholar]
- 30.Agnafors S, Barmark M, Sydsjö G. Mental health and academic performance: a study on selection and causation effects from childhood to early adulthood. Soc Psychiatry Psychiatr Epidemiol. 2021;56(5):857-866. doi: 10.1007/s00127-020-01934-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Eisenberg D, Golberstein E, Hunt JB. Mental health and academic success in college. BE J Econ Anal Policy. 2009;9(1):40. doi: 10.2202/1935-1682.2191 [DOI] [Google Scholar]
- 32.Zhang J, Zheng Y. How do academic stress and leisure activities influence college students’ emotional well-being? a daily diary investigation. J Adolesc. 2017;60:114-118. doi: 10.1016/j.adolescence.2017.08.003 [DOI] [PubMed] [Google Scholar]
- 33.Melgaard J, Monir R, Lasrado LA, Fagerstrøm A. Academic procrastination and online learning during the COVID-19 pandemic. Procedia Comput Sci. 2022;196:117-124. doi: 10.1016/j.procs.2021.11.080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Bettinger EP, Fox L, Loeb S, Taylor ES. Virtual classrooms: how online college courses affect student success. Am Econ Rev. 2017;107(9):2855-2875. doi: 10.1257/aer.20151193 [DOI] [Google Scholar]
- 35.Tasso AF, Hisli Sahin N, San Roman GJ. COVID-19 disruption on college students: academic and socioemotional implications. Psychol Trauma. 2021;13(1):9-15. doi: 10.1037/tra0000996 [DOI] [PubMed] [Google Scholar]
- 36.Reyes JF. More New Jersey schools are going remote for the beginning of 2022. The Philadelphia Inquirer. Published January 1, 2022. Accessed February 18, 2022. https://www.inquirer.com/education/new-jersey-schools-remote-learning-20220101.html
- 37.Silva D, Przybyla H. Some U.S. schools switch to remote learning, delay start of classes as Omicron surge disrupts return from winter break: large school districts in Georgia, Michigan, New Jersey, Ohio and Wisconsin are among those whose plans have changed abruptly. NBC News. Published January 3, 2022. Accessed February 18, 2022. https://www.nbcnews.com/news/us-news/us-schools-switch-remote-learning-delay-start-classes-omicron-surge-di-rcna10795
- 38.Kim MC, Kweon OJ, Lim YK, Choi SH, Chung JW, Lee MK. Impact of social distancing on the spread of common respiratory viruses during the coronavirus disease outbreak. PLoS One. 2021;16(6):e0252963. doi: 10.1371/journal.pone.0252963 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Davitt ED, Heer MM, Winham DM, Knoblauch ST, Shelley MC. Effects of COVID-19 on university student food security. Nutrients. 2021;13(6):1932. doi: 10.3390/nu13061932 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Jaffe AE, Kumar SA, Ramirez JJ, DiLillo D. Is the COVID-19 pandemic a high-risk period for college student alcohol use? a comparison of three spring semesters. Alcohol Clin Exp Res. 2021;45(4):854-863. doi: 10.1111/acer.14572 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Lim LTS, Regencia ZJG, Dela Cruz JRC, et al. Assessing the effect of the COVID-19 pandemic, shift to online learning, and social media use on the mental health of college students in the Philippines: a mixed-method study protocol. PLoS One. 2022;17(5):e0267555. doi: 10.1371/journal.pone.0267555 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eAppendix. Covariates
eReferences.