Table 3.
Regression estimates of percentage point changes in mental health problems from 2017–2019 to April–June 2020 comparing differences between population subgroups
| Variable | Subgroup differences in changes in mental health from 2017–2019 to April 2020 | Subgroup differences in changes in mental health recovery from April to June 2020 | ||
|---|---|---|---|---|
| (%)a | 95% CI | (%)a | 95% CI | |
| Age group (comparison is 50–64 years) | ||||
| 18–34 years | +9.3*** | (4.2–14.4) | −6.7** | (−11.8 to −1.6) |
| 35–49 years | +5.3** | (1.5–9.2) | −2.7 | (−6.5 to 1.1) |
| 65+ years | +3.1 | (−1.1 to 7.3) | −2.6 | (−6.7 to 1.5) |
| Femaleb | +6.1*** | (3.0–9.3) | −4.6** | (−7.8 to −1.3) |
| Whitec | +0.8 | (−6.2 to 7.9) | −2.5 | (−8.6 to 3.6) |
| Marriedd | 0.0 | (−3.4 to 3.4) | −0.5 | (−3.7 to 2.8) |
| University degreee | +5.7*** | (2.7–8.8) | −2.6 | (−5.7 to 0.5) |
| Income levelf (comparison is low) | ||||
| Middle tertile | +1.7 | (−2.6 to 6.0) | −2.2 | (−6.3 to 1.9) |
| Top tertile | +5.6** | (1.8–9.5) | −3.0 | (−6.7 to 0.8) |
| COVID-19 risk | +6.5 | (−2.3 to 15.3) | −2.5 | (−9.7 to 4.7) |
Note: Estimates are from marginal effects calculated after a logistic regression with standard errors adjusted for clustering at the individual-level and controlling for all characteristics presented. Age groups are based on age reported in the COVID-19 surveys.
Those with a GHQ ‘caseness’ score ⩾3 were classified as experiencing mental health problems.
Difference between females and males in the change in mental health problems between time points.
Difference between whites and non-whites in the change in mental health problems between time points.
Difference between married and non-married participants in the change in mental health problems between time points.
Difference between those with/without a degree in the change in mental health problems between time points.
Net household income in the 2017–2019 wave of the UKHLS.
*p < 0.05; **p < 0.01; ***p < 0.001.