Table 2.
Sub-group analyses for standardized mean change (SMC) in mental health symptoms.
| Moderator | Test of sub-group difference | Sub-groups | Effect estimate |
|---|---|---|---|
| Age | QM(1) = 0.001, p = .978, I2 = 97.9% | Adults (n = 127) | SMC = .105 [95% CI: .033 to .178], z = 2.84, p = .005, I2 = 98.0% |
| Child/adolescents (n = 38) | SMC = .114 [95% CI: -.030 to .257], z = 1.56, p = .120, I2 = 94.6%% | ||
| Gender | QM(1) = 2.73, p = .098, I2 = 93.8% | Males (n = 13) | SMC = .086 [95% CI: -.030 to .202], z = .145, p = .147, I2 = 89.9% |
| Females (n = 14) | SMC = .150 [95% CI: .005 to .295], z = 2.02, p = .043, I2 = 95.0% | ||
| Continent | QM(3) = 0.916, p = .822, I2 = 97.9% | Europe (n = 100) | SMC = .093 [95% CI: .002 to .185], z = 2.00, p = .046, I2 = 97.8% |
| North America (n = 45) | SMC = .132 [95% CI: -.012 to .277], z = 1.79, p = .073, I2 = 98.5% | ||
| Other (n = 15) | SMC = .160 [95% CI: .085 to .234], z = 4.20, p < .001, I2 = 87.3% | ||
| Mix (n = 5) | SMC = .052 [95% CI: -.081 to .184], z = 0.76, p = .446, I2 = 72.1% | ||
| Population | QM(3) = 8.735, p = .033, I2 = 97.8% | General population (n = 75) | SMC = .118 [95% CI: .042 to .193], z = 3.04, p = .002, I2 = 97.9% |
| University student (n = 40) | SMC = .133 [95% CI: -.005 to .272], z = 1.87, p = p = .059, I2 = 95.2% | ||
| Pre-existing mental health condition (n = 25) | SMC = -.017 [95% CI: -.211 to .178], z = 0.17, = .867, I2 = 97.7% | ||
| Pre-existing physical health condition (n = 14) | SMC = .249 [95% CI: .067 to .431], z = 2.68, p = .007, I2 = 97.9% | ||
| Publication status | QM(1) = 0.053, p = .818, I2 = 97.9% | Journal article (n = 122) | SMC = .101 [95% CI: .021 to .181], z = 2.48, p = .013, I2 = 98.1% |
| Pre-print (n = n = 43) | SMC = .121 [95% CI: .001 to .241], z = 1.97, p = .049, I2 = 96.5% | ||
| Representative sampling | QM(1) = 0.114, p = .735, I2 = 97.9% | Reported (n = 34) | SMC = .125 [95% CI: .043 to .206], z = 3.01, p = p = .003, I2 = 98.1% |
| Not reported or unclear (n = 131) | SMC = .101 [95% CI: .020 to .181], z = 2.43, p = .015, I2 = 96.2% | ||
| Sample size | QM(1) = 0.047, p = .829, I2 = 97.9% | ≥1000 (n = n = 24) | SMC = .108 [95% CI: .013 to .203], z = 2.32, p = .013, I2 = 98.9% |
| <1000 (n = 141) | SMC = .108 [95% CI: .031 to .184], z = 2.77, p = .006, I2 = 94.6% | ||
| Pre-pandemic measurea | QM(1) = 0.167, p = .683, I2 = 97.9% | More than 12 months prior to pandemic measure (n = 42) | SMC = .092 [95% CI: -.008 to .193], z = 1.80, p = p = .072, I2 = 98.3% |
| 12 months or less prior to pandemic measure (n = 119) | SMC = .117 [95% CI: .032 to .203], z = 2.70, p = .007, I2 = 97.3% | ||
| Mode of survey | QM(1) = 5.056, p = .025, I2 = 97.9% | Consistent between pre and during pandemic (n = 82) | SMC = .179 [95% CI: .098 to .260], z = 4.31, p < .001, I2 = 98.1% |
| Inconsistent or unclear (n = 83) | SMC = .034 [95% CI: -.064 to .132], z = 0.685, p = .493, I2 = 95.9% | ||
| Conflicts of interest | QM(1) = 0.111, p = .739, I2 = 97.9% | Statement included + no reported conflicts (n = 124) | SMC = .112 [95% CI: .041 to .184], z = 3.07, p =.001, I2 = 97.6% |
| No conflict statement or reported conflicts (n = 41) | SMC = .089 [95% CI: -.065 to .243], z = 1.32, p = .257, I2 = 98.1% | ||
| Attritionb | QM(1) = 0.045, p = p = .832, I2 = 97.8% | Level of attrition ≤25% (n = 47) | SMC = .038 [95% CI: -.055 to .131], z = 0.80, p = .424, I2 = 98.5% |
| Level of attrition > 25% (n = 66) | SMC = .103 [95% CI: -.002 to .208], z = .1.93, p = .054, I2 = 95.4% |
a
Number of comparisons does not add up to n = 165 because for a small number of effects it was unclear whether the pre-pandemic measures was collected more than 12 months prior to the post-pandemic outbreak measure.
b
We also examined whether attrition was associated with the change in mental health symptoms pre-post pandemic using meta-regression. There was no significant association (coefficient = .092 [95% CI: -.185 to .370] z = .653, p = .514).