Associations between COVID-19 risk perceptions and mental health, wellbeing, and risk behaviours

Figures & data

Figure 1. Cross-sectional associations between COVID-19 holistic risk perceptions and mental health, wellbeing, and risk behaviours.

Note. Whole sample. Forest plot shows the fully adjusted odds ratios (circles) and 95% confidence intervals (bars). Fully adjusted = adjusted for age, gender, education, keyworker status, pre-pandemic anxiety, depression, high-risk drinking, smoking, and early pandemic suspected COVID-19 infection.

Table 1. Cross-sectional associations between COVID-19 risk perceptions and mental health, wellbeing, and risk behaviours (whole sample).

	COVID-19 risk perceptions (Exposures)
	Holistic			Cognitive			Affective			Self			Other
Outcome and model	OR (95% CI)	P	N	OR (95% CI)	P	N	OR (95% CI)	P	N	OR (95% CI)	P	N	OR (95% CI)	P	N
Generalised anxiety disorder
Unadjusted	2.75 (2.35, 3.21)	<.001	4982	1.60 (1.38, 1.86)	<.001	5065	2.74 (2.34, 3.22)	<.001	4994	2.16 (1.86, 2.52)	<.001	5038	3.00 (2.55, 3.53)	<.001	5011
Partially adjusted	3.05 (2.55, 3.66)	<.001	4244	1.52 (1.28, 1.80)	<.001	4314	3.12 (2.58, 3.77)	<.001	4255	2.78 (2.32, 3.33)	<.001	4289	2.97 (2.46, 3.58)	<.001	4270
Fully adjusted	2.78 (2.20, 3.52)	<.001	2780	1.50 (1.19, 1.89)	.001	2820	2.70 (2.12, 3.44)	<.001	2787	2.63 (2.08, 3.32)	<.001	2806	2.71 (2.12, 3.45)	<.001	2796
Depression
Unadjusted	1.81 (1.52, 2.16)	<.001	4922	1.42 (1.20, 1.69)	<.001	5006	1.81 (1.51, 2.16)	<.001	4933	1.45 (1.22, 1.72)	<.001	4977	2.03 (1.70, 2.43)	<.001	4951
Partially adjusted	1.90 (1.55, 2.34)	<.001	4198	1.23 (1.00, 1.50)	.048	4268	1.94 (1.57, 2.41)	<.001	4208	1.83 (1.49, 2.26)	<.001	4242	2.02 (1.63, 2.49)	<.001	4224
Fully adjusted	1.65 (1.24, 2.18)	<.001	2753	1.12 (0.85, 1.47)	.430	2793	1.77 (1.33, 2.36)	<.001	2760	1.57 (1.19, 2.09)	.002	2778	1.92 (1.44, 2.56)	<.001	2770
Low wellbeing
Unadjusted	1.84 (1.62, 2.09)	<.001	4942	1.32 (1.16, 1.50)	<.001	5025	1.84 (1.62, 2.10)	<.001	4954	1.57 (1.38, 1.78)	<.001	4997	1.77 (1.56, 2.01)	<.001	4972
Partially adjusted	1.91 (1.65, 2.21)	<.001	4213	1.29 (1.11, 1.49)	.001	4283	1.94 (1.67, 2.26)	<.001	4224	1.82 (1.57, 2.12)	<.001	4257	1.74 (1.50, 2.01)	<.001	4240
Fully adjusted	1.76 (1.45, 2.13)	<.001	2766	1.18 (0.97, 1.43)	.090	2806	1.83 (1.50, 2.22)	<.001	2773	1.79 (1.48, 2.17)	<.001	2791	1.56 (1.29, 1.89)	<.001	2783
High-risk drinking
Unadjusted	0.81 (0.72, 0.91)	<.001	5022	1.14 (1.01, 1.27)	.027	5107	0.76 (0.67, 0.85)	<.001	5034	0.75 (0.67, 0.84)	<.001	5077	0.91 (0.81, 1.02)	.095	5051
Partially adjusted	0.88 (0.78, 1.01)	.061	4287	1.09 (0.97, 1.24)	.154	4358	0.84 (0.74, 0.96)	.009	4297	0.87 (0.77, 0.99)	.039	4331	0.95 (0.84, 1.08)	.458	4313
Fully adjusted	0.95 (0.79, 1.13)	.537	2809	1.18 (0.99, 1.40)	.059	2849	0.89 (0.74, 1.06)	.188	2816	0.98 (0.82, 1.18)	.860	2834	0.97 (0.82, 1.15)	.726	2826
Increased alcohol use
Unadjusted	1.31 (1.16, 1.48)	<.001	4334	1.11 (0.98, 1.25)	.092	4405	1.31 (1.16, 1.47)	<.001	4343	1.22 (1.09, 1.38)	.001	4379	1.22 (1.09, 1.38)	.001	4357
Partially adjusted	1.40 (1.23, 1.60)	<.001	3698	1.13 (0.99, 1.29)	.069	3758	1.39 (1.22, 1.59)	<.001	3704	1.31 (1.15, 1.50)	<.001	3734	1.28 (1.12, 1.46)	<.001	3719
Fully adjusted	1.46 (1.24, 1.72)	<.001	2541	1.13 (0.96, 1.32)	.140	2578	1.52 (1.29, 1.79)	<.001	2544	1.39 (1.18, 1.64)	<.001	2561	1.29 (1.10, 1.52)	.001	2556
Increased smoking/e-cigarette use
Unadjusted	1.41 (1.08, 1.82)	.010	941	1.13 (0.87, 1.46)	.369	963	1.30 (1.00, 1.68)	.050	942	1.14 (0.88, 1.48)	.307	957	1.47 (1.13, 1.90)	.004	946
Partially adjusted	1.45 (1.06, 1.98)	.019	726	1.21 (0.89, 1.63)	.225	741	1.26 (0.92, 1.73)	.146	726	1.24 (0.90, 1.69)	.188	736	1.45 (1.07, 1.97)	.017	730
Fully adjusted	1.14 (0.72, 1.80)	.586	420	0.97 (0.63, 1.49)	.888	426	0.98 (0.63, 1.54)	.942	420	1.00 (0.64, 1.57)	.989	422	1.31 (0.85, 2.03)	.222	423
Self-isolating given suspected COVID-19 infection
Unadjusted	1.40 (1.04, 1.90)	.028	758	2.24 (1.65, 3.05)	<.001	777	1.09 (0.80, 1.47)	.589	761	0.83 (0.61, 1.12)	.218	765	1.35 (1.00, 1.81)	.047	769
Partially adjusted	1.60 (1.14, 2.26)	.007	638	2.40 (1.70, 3.37)	<.001	655	1.27 (0.90, 1.78)	.177	640	0.92 (0.65, 1.30)	.634	643	1.40 (1.01, 1.95)	.044	649
Fully adjusted	1.74 (1.13, 2.68)	.012	413	2.27 (1.48, 3.48)	<.001	422	1.30 (0.85, 2.01)	.231	415	0.98 (0.63, 1.54)	.943	416	1.42 (0.94, 2.16)	.096	421
Face-to-face contact outside household
Unadjusted	0.86 (0.77, 0.97)	.011	5029	0.86 (0.77, 0.97)	.010	5115	0.86 (0.77, 0.97)	.012	5042	1.00 (0.90, 1.13)	.946	5085	0.86 (0.77, 0.96)	.010	5059
Partially adjusted	0.77 (0.67, 0.88)	<.001	4293	0.91 (0.80, 1.04)	.160	4365	0.74 (0.65, 0.85)	<.001	4304	0.83 (0.72, 0.94)	.005	4338	0.83 (0.73, 0.95)	.006	4320
Fully adjusted	0.83 (0.70, 0.98)	.027	2816	0.91 (0.77, 1.08)	.280	2857	0.78 (0.66, 0.92)	.004	2823	0.82 (0.69, 0.97)	.019	2842	0.88 (0.75, 1.05)	.149	2833
Physical contact outside household
Unadjusted	0.84 (0.74, 0.97)	.016	4733	0.79 (0.69, 0.91)	.001	4812	0.84 (0.74, 0.96)	.013	4744	0.84 (0.73, 0.96)	.011	4785	0.82 (0.71, 0.94)	.004	4761
Partially adjusted	0.85 (0.73, 0.98)	.030	4046	0.82 (0.71, 0.95)	.008	4112	0.83 (0.72, 0.97)	.017	4056	0.81 (0.70, 0.94)	.007	4087	0.83 (0.71, 0.96)	.012	4072
Fully adjusted	0.83 (0.68, 1.00)	.049	2662	0.78 (0.65, 0.94)	.008	2700	0.79 (0.65, 0.95)	.013	2668	0.71 (0.60, 0.86)	<.001	2685	0.84 (0.69, 1.01)	.057	2678

Note. Logistic regressions. OR = odds ratio. CI = confidence interval. Partially adjusted = adjusted for sociodemographic variables (age, gender, education, and keyworker status). Fully adjusted = additionally adjusted for prior mental health and risk behaviour variables (anxiety, depression, high-risk drinking, smoking, and suspected COVID-19 infection). All variables in the models are binary. All risk perception variables were dichotomised at the median.

Table 2. Longitudinal associations between pre-pandemic mental health, wellbeing, and risk behaviours and COVID-19 risk perceptions (whole sample).

	COVID-19 risk perceptions (Outcomes)
	Holistic			Cognitive			Affective			Self			Other
Exposure and model	OR (95% CI)	P	N	OR (95% CI)	P	N	OR (95% CI)	P	N	OR (95% CI)	P	N	OR (95% CI)	P	N
Pre-pandemic anxiety
Unadjusted	1.52 (1.30, 1.78)	<.001	4165	1.09 (0.93, 1.28)	.276	4235	1.65 (1.41, 1.93)	<.001	4175	1.52 (1.30, 1.78)	<.001	4207	1.41 (1.20, 1.65)	<.001	4192
Partially adjusted	1.47 (1.23, 1.75)	<.001	3673	1.24 (1.04, 1.47)	.014	3733	1.54 (1.29, 1.84)	<.001	3682	1.25 (1.05, 1.49)	.012	3709	1.50 (1.27, 1.79)	<.001	3697
Fully adjusted	1.64 (1.29, 2.09)	<.001	2533	1.34 (1.06, 1.71)	.016	2570	1.74 (1.37, 2.22)	<.001	2539	1.44 (1.13, 1.84)	.003	2557	1.75 (1.37, 2.22)	<.001	2547
Pre-pandemic depression
Unadjusted	1.30 (1.11, 1.53)	.002	4174	1.02 (0.87, 1.20)	.815	4243	1.34 (1.14, 1.58)	<.001	4184	1.21 (1.03, 1.42)	.022	4216	1.22 (1.04, 1.44)	.016	4201
Partially adjusted	1.17 (0.98, 1.41)	.084	3680	1.10 (0.92, 1.31)	.294	3740	1.23 (1.02, 1.47)	.026	3689	1.02 (0.85, 1.23)	.808	3716	1.18 (0.99, 1.41)	.069	3705
Fully adjusted	0.94 (0.73, 1.22)	.648	2533	0.98 (0.77, 1.26)	.903	2570	0.95 (0.74, 1.23)	.716	2539	0.86 (0.66, 1.12)	.263	2557	0.88 (0.68, 1.14)	.341	2547
Pre-pandemic low wellbeing
Unadjusted	1.51 (1.30, 1.75)	<.001	4056	1.19 (1.03, 1.38)	.019	4125	1.58 (1.36, 1.84)	<.001	4067	1.55 (1.34, 1.80)	<.001	4102	1.45 (1.25, 1.69)	<.001	4080
Partially adjusted	1.47 (1.25, 1.74)	<.001	3597	1.24 (1.06, 1.45)	.008	3657	1.52 (1.29, 1.80)	<.001	3606	1.40 (1.19, 1.65)	<.001	3636	1.46 (1.24, 1.72)	<.001	3619
Fully adjusted	1.41 (1.15, 1.74)	.001	2465	1.19 (0.97, 1.46)	.101	2502	1.52 (1.24, 1.87)	<.001	2471	1.53 (1.24, 1.89)	<.001	2489	1.34 (1.09, 1.65)	.005	2479
Pre-pandemic high-risk drinking
Unadjusted	0.79 (0.69, 0.90)	<.001	3738	1.06 (0.93, 1.20)	.388	3796	0.73 (0.64, 0.84)	<.001	3748	0.63 (0.56, 0.72)	<.001	3777	0.93 (0.82, 1.06)	.290	3760
Partially adjusted	0.90 (0.76, 1.04)	.144	3332	0.94 (0.81, 1.08)	.371	3382	0.88 (0.76, 1.02)	.079	3341	0.80 (0.69, 0.93)	.003	3365	0.98 (0.85, 1.13)	.757	3352
Fully adjusted	0.89 (0.75, 1.06)	.192	2533	0.97 (0.82, 1.14)	.702	2570	0.89 (0.74, 1.05)	.169	2539	0.78 (0.65, 0.92)	.004	2557	1.02 (0.86, 1.21)	.807	2547
Pre-pandemic smoking
Unadjusted	1.05 (0.89, 1.24)	.573	4135	1.06 (0.90, 1.26)	.485	4198	1.04 (0.87, 1.23)	.687	4145	0.93 (0.78, 1.10)	.384	4175	1.28 (1.08, 1.52)	.005	4159
Partially adjusted	1.11 (0.91, 1.35)	.308	3670	1.01 (0.83, 1.22)	.957	3726	1.10 (0.91, 1.34)	.332	3679	1.14 (0.94, 1.38)	.186	3705	1.22 (1.00, 1.48)	.048	3692
Fully adjusted	0.96 (0.74, 1.24)	.755	2533	1.01 (0.78, 1.29)	.966	2570	0.94 (0.73, 1.22)	.667	2539	1.08 (0.84, 1.40)	.529	2557	1.12 (0.86, 1.45)	.394	2547

Note. Logistic regressions. OR = odds ratio. CI = confidence interval. Partially adjusted = adjusted for sociodemographic variables (age, gender, education, and keyworker status). Fully adjusted = additionally adjusted for prior mental health and risk behaviour variables (pre-pandemic anxiety, depression, high-risk drinking, smoking, and early pandemic suspected COVID-19 infection). The same sociodemographic variables are included in all partially adjusted models. However, the variables in the fully adjusted models differ based on the exposure in each model (e.g. pre-pandemic anxiety is removed as a confounder when pre-pandemic anxiety is the exposure). All variables in the models are binary. All risk perception variables were dichotomised at the median.

Figure 2. Longitudinal associations between pre-pandemic and early pandemic variables and COVID-19 holistic risk perceptions.

Note. Whole sample. Forest plot shows the fully adjusted odds ratios (circles) and 95% confidence intervals (bar). Fully adjusted = adjusted for age, gender, education, keyworker status, pre-pandemic anxiety, depression, high-risk drinking, smoking, and early pandemic suspected COVID-19 infection.

Table 3. Longitudinal associations between early pandemic risk behaviours and COVID-19 risk perceptions (whole sample).

	COVID-19 risk perceptions (Outcomes)
	Holistic			Cognitive			Affective			Self			Other
Exposure and model	OR (95% CI)	P	N	OR (95% CI)	P	N	OR (95% CI)	P	N	OR (95% CI)	P	N	OR (95% CI)	P	N
Early pandemic self-isolating given suspected COVID-19 infection
Unadjusted	1.45 (0.88, 2.37)	.142	394	1.27 (0.79, 2.05)	.323	410	1.57 (0.96, 2.57)	.073	395	1.29 (0.79, 2.11)	.304	400	1.33 (0.82, 2.15)	.245	401
Partially adjusted	1.11 (0.63, 1.94)	.724	352	1.41 (0.82, 2.44)	.212	366	1.20 (0.68, 2.10)	.534	353	0.92 (0.52, 1.62)	.766	357	1.05 (0.60, 1.82)	.868	359
Fully adjusted	1.26 (0.64, 2.48)	.500	245	1.53 (0.81, 2.87)	.188	253	1.44 (0.73, 2.86)	.295	233	0.83 (0.42, 1.64)	.587	248	1.31 (0.69, 2.47)	.413	251
Early pandemic face-to-face contact outside household
Unadjusted	0.95 (0.84, 1.08)	.430	4056	0.80 (0.71, 0.91)	.001	4123	0.97 (0.86, 1.10)	.629	4067	1.01 (0.89, 1.14)	.890	4097	0.91 (0.80, 1.03)	.133	4081
Partially adjusted	0.89 (0.77, 1.02)	.096	3545	0.83 (0.72, 0.95)	.008	3604	0.92 (0.80, 1.06)	.270	3553	0.90 (0.78, 1.04)	.151	3581	0.93 (0.81, 1.07)	.289	3567
Fully adjusted	0.93 (0.78, 1.11)	.434	2411	0.88 (0.75, 1.05)	.155	2446	0.98 (0.82, 1.16)	.805	2415	0.92 (0.77, 1.10)	.355	2432	1.05 (0.88, 1.24)	.609	2425
Early pandemic physical contact outside household
Unadjusted	0.86 (0.72, 1.03)	.103	3619	0.82 (0.68, 0.98)	.026	3674	0.93 (0.78, 1.11)	.411	3628	0.93 (0.78, 1.11)	.439	3653	0.89 (0.75, 1.07)	.220	3641
Partially adjusted	0.89 (0.73, 1.08)	.243	3166	0.81 (0.66, 0.98)	.030	3213	0.97 (0.80, 1.18)	.775	3174	0.98 (0.81, 1.19)	.835	3195	0.92 (0.76, 1.12)	.401	3185
Fully adjusted	0.93 (0.73, 1.19)	.563	2159	0.91 (0.71, 1.15)	.418	2189	0.94 (0.74, 1.19)	.596	2163	1.01 (0.79, 1.28)	.965	2177	1.08 (0.85, 1.37)	.539	2172

Note. Logistic regressions. OR = odds ratio. CI = confidence interval. Partially adjusted = adjusted for sociodemographic variables (age, gender, education, and keyworker status). Fully adjusted = additionally adjusted for prior mental health and risk behaviour variables (pre-pandemic anxiety, depression, high-risk drinking, smoking, and early pandemic suspected COVID-19 infection). All variables in the models are binary. All risk perception variables were dichotomised at the median.

Table 4. Prospective longitudinal associations between mental health and wellbeing polygenic risk scores and COVID-19 risk perceptions.

	COVID-19 risk perceptions (Outcomes)
	Holistic			Cognitive			Affective			Self			Other
Exposure and model	b (95% CI)	P	N	b (95% CI)	P	N	b (95% CI)	P	N	b (95% CI)	P	N	b (95% CI)	P	N
Whole sample
Anxiety
Unadjusted	0.12 (−0.07, 0.32)	.206	3615	0.03 (−0.03, 0.09)	.407	3672	0.09 (−0.08, 0.25)	.308	3623	0.04 (−0.06, 0.13)	.430	3652	0.08 (−0.03, 0.20)	.166	3633
Fully adjusted	0.12 (−0.08, 0.31)	.236	3615	0.02 (−0.04, 0.08)	.465	3672	0.08 (−0.08, 0.25)	.331	3623	0.03 (−0.06, 0.13)	.471	3652	0.08 (−0.04, 0.20)	.193	3633
Depression
Unadjusted	0.22 (0.04, 0.41)	.018	3615	0.02 (−0.04, 0.08)	.578	3672	0.20 (0.04, 0.36)	.015	3623	0.08 (−0.01, 0.17)	.077	3652	0.14 (0.03, 0.26)	.014	3633
Fully adjusted	0.21 (0.02, 0.40)	.029	3615	0.01 (−0.05, 0.07)	.663	3672	0.19 (0.03, 0.35)	.023	3623	0.06 (−0.03, 0.15)	.171	3652	0.15 (0.03, 0.26)	.013	3633
Wellbeing
Unadjusted	−0.29 (−0.49, −0.10)	.003	3615	−0.02 (−0.08, 0.04)	.501	3672	−0.27 (−0.44, −0.10)	.001	3623	−0.12 (−0.21, −0.03)	.007	3652	−0.17 (−0.29, −0.05)	.006	3633
Fully adjusted	−0.29 (−0.48, −0.09)	.004	3615	−0.01 (−0.08, 0.05)	.630	3672	−0.27 (−0.44, −0.10)	.002	3623	−0.11 (−0.21, −0.02)	.013	3652	−0.17 (−0.29, −0.05)	.006	3633
Mother sample
Anxiety
Unadjusted	0.11 (−0.16, 0.38)	.426	1792	−0.01 (−0.09, 0.07)	.838	1824	0.12 (−0.12, 0.35)	.334	1796	0.05 (−0.07, 0.18)	.400	1806	0.06 (−0.11, 0.22)	.513	1807
Fully adjusted	0.11 (−0.17, 0.38)	.444	1792	−0.01 (−0.09, 0.07)	.810	1824	0.12 (−0.12, 0.35)	.338	1796	0.05 (−0.08, 0.18)	.444	1806	0.06 (−0.11, 0.22)	.507	1807
Depression
Unadjusted	0.26 (−0.01, 0.53)	.057	1792	0.04 (−0.04, 0.12)	.366	1824	0.23 (−0.00, 0.46)	.054	1796	0.10 (−0.03, 0.22)	.134	1806	0.17 (0.01, 0.34)	.037	1807
Fully adjusted	0.26 (−0.02, 0.53)	.067	1792	0.03 (−0.05, 0.11)	.495	1824	0.24 (−0.00, 0.47)	.052	1796	0.08 (−0.05, 0.21)	.209	1806	0.19 (0.02, 0.36)	.027	1807
Wellbeing
Unadjusted	−0.33 (−0.60, −0.05)	.020	1792	−0.03 (−0.11, 0.05)	.470	1824	−0.31 (−0.54, −0.07)	.011	1796	−0.14 (−0.27, −0.01)	.032	1806	−0.19 (−0.36, −0.02)	.026	1807
Fully adjusted	−0.32 (−0.60, −0.03)	.030	1792	−0.02 (−0.11, 0.07)	.675	1824	−0.31 (−0.56, −0.06)	.014	1796	−0.13 (−0.26, 0.01)	.065	1806	−0.20 (−0.37, −0.02)	.027	1807
Young person sample
Anxiety
Unadjusted	0.18 (−0.07, 0.42)	.158	1823	0.05 (−0.04, 0.13)	.274	1848	0.11 (−0.10, 0.32)	.305	1827	0.06 (−0.05, 0.18)	.293	1846	0.11 (−0.05, 0.26)	.178	1826
Fully adjusted	0.15 (−0.10, 0.40)	.252	1823	0.04 (−0.04, 0.13)	.318	1848	0.08 (−0.13, 0.29)	.455	1827	0.05 (−0.07, 0.17)	.386	1846	0.09 (−0.07, 0.25)	.283	1826
Depression
Unadjusted	0.20 (−0.05, 0.46)	.113	1823	−0.01 (−0.09, 0.07)	.812	1848	0.19 (−0.02, 0.41)	.079	1827	0.08 (−0.04, 0.19)	.213	1846	0.11 (−0.05, 0.27)	.173	1826
Fully adjusted	0.16 (−0.10, 0.42)	.218	1823	−0.02 (−0.10, 0.07)	.706	1848	0.16 (−0.06, 0.38)	.162	1827	0.06 (−0.07, 0.18)	.364	1846	0.09 (−0.07, 0.25)	.285	1826
Wellbeing
Unadjusted	−0.29 (−0.54, −0.04)	.025	1823	−0.00 (−0.08, 0.08)	.979	1848	−0.27 (−0.48, −0.06)	.013	1827	−0.13 (−0.25, −0.02)	.025	1846	−0.14 (−0.30, 0.02)	.077	1826
Fully adjusted	−0.26 (−0.53, −0.00)	.048	1823	0.00 (−0.09, 0.09)	.993	1848	−0.24 (−0.47, −0.02)	.031	1827	−0.12 (−0.25, −0.00)	.049	1846	−0.13 (−0.30, 0.03)	.121	1826

Note. Linear regressions. b = unstandardised beta coefficient. CI = confidence interval. Polygenic risk scores were created using the p-value threshold of 0.05. Fully adjusted models = adjusted for genetic principal components of ancestry (PC1-PC10).

Data availability statement

The analysis code is available from the University of Bristol’s Research Data Repository, data.bris, at https://doi.org/10.5523/bris.34bmhh800n6pb25aeva7qucjre. GWAS summary statistics used to create the PRS are available from the original publications. The informed consent obtained from ALSPAC participants does not allow the data to be made freely available through any third party maintained public repository. However, data used for this submission can be made available on request to the ALSPAC Executive. The ALSPAC data management plan describes in detail the policy regarding data sharing, which is through a system of managed open access. Full instructions for applying for data access can be found here: http://www.bristol.ac.uk/alspac/researchers/access/. The ALSPAC study website contains details of all the data that are available (http://www.bristol.ac.uk/alspac/researchers/our-data/).