Skip to main content

Table 3 Odds ratios (95% CIs) of univariable associations between potential predictors and long COVID 3 months after a PCR test, overall and stratified by SARS-CoV-2 status

From: Predictive model for long COVID in children 3 months after a SARS-CoV-2 PCR test

Potential predictor

Total population (N=7139)

SARS-CoV-2 negative (N=3893)

SARS-CoV-2 positive (N=3246)

Pinteractiona

Odds ratio (95% CIs)

p value

Odds ratio (95% CIs)

p value

Odds ratio (95% CIs)

p value

SARS-CoV-2 status

 Negative

Ref

<0.001

     

 Positive

1.48 (1.33, 1.66)

 

N/A

 

N/A

 

N/A

Sex

 Male

Ref

<0.001

Ref

<0.001

Ref

<0.001

0.5858

 Female

2.02 (1.78, 2.30)

 

2.11 (1.75, 2.54)

 

1.96 (1.65, 2.34)

  

Age (years)b

 11–13

Ref

<0.001

Ref

<0.001

Ref

<0.001

0.2533

 14–15

1.54 (1.32, 1.80)

 

1.71 (1.37, 2.13)

 

1.36 (1.09, 1.70)

  

 16–17

1.73 (1.50, 1.99)

1.73 (1.42, 2.12)

1.70 (1.40, 2.07)

Index of Multiple Deprivation

 Quintile 1 (most deprived)

Ref

<0.001

Ref

0.034

Ref

0.004

0.149

 Quintile 2

0.99 (0.83, 1.17)

 

0.82 (0.64, 1.05)

 

1.18 (0.93, 1.49)

  

 Quintile 3

0.88 (0.74, 1.05)

0.79 (0.62, 1.02)

1.00 (0.78, 1.28)

 Quintile 4

0.79 (0.66, 0.94)

0.69 (0.53, 0.89)

0.90 (0.70, 1.16)

 Quintile 5 (least deprived)

0.72 (0.60, 0.86)

0.72 (0.56, 0.93)

0.72 (0.56, 0.93)

Ethnicity

 White

Ref

0.055

Ref

0.146

Ref

0.131

0.417

 Asian/Asian British

0.93 (0.79, 1.09)

 

0.93 (0.73, 1.18)

 

0.90 (0.72, 1.13)

  

 Black/African/Caribbean

1.11 (0.83, 1.48)

1.36 (0.92, 2.01)

0.89 (0.57, 1.38)

 Mixed

1.38 (1.09, 1.76)

1.43 (1.03, 2.00)

1.36 (0.95, 1.93)

 Others

0.70 (0.43, 1.16)

1.00 (0.50, 2.00)

0.49 (0.24, 1.00)

 Preferred not to say

0.96 (0.49, 1.87)

0.59 (0.18, 1.97)

1.23 (0.54, 2.83)

Self-rated physical health

 Very good

Ref

<0.001

Ref

<0.001

Ref

<0.001

<0.001

 Good

2.06 (1.77, 2.39)

 

2.49 (1.98, 3.13)

 

1.80 (1.47, 2.20)

  

 Okay

3.57 (3.04, 4.20)

4.51 (3.54, 5.76)

3.00 (2.40, 3.74)

 Poor/very poor

7.60 (5.41, 10.68)

14.91 (9.32, 23.85)

3.71 (2.23, 6.19)

Self-rated mental health

 Very good

Ref

<0.001

Ref

<0.001

Ref

<0.001

0.015

 Good

2.12 (1.75, 2.56)

 

2.55 (1.87, 3.81)

 

1.92 (1.50, 2.46)

  

 Okay

3.91 (3.23, 4.73)

5.50 (4.06, 7.46)

3.06 (2.39, 3.93)

 Poor/very poor

12.72 (10.17, 15,91)

17.46 (12.43, 24.53)

10.54 (7.72, 14.40)

Lonelinessc

 Never

Ref

<0.001

Ref

<0.001

Ref

<0.001

0.002

 Hardly ever

1.75 (1.45, 2.11)

 

2.16 (1.59, 2.93)

 

1.54 (1.21, 1.96)

  

 Occasionally

3.91 (3.23, 4.73)

5.28 (3.91, 7.14)

3.30 (2.56, 4.26)

 Some of the time

4.81 (3.98, 5.80)

6.86 (5.10, 9.22)

3.82 (2.97, 4.91)

 Often/always

7.98 (6.34, 10.04)

13.41 (9.56, 18.80)

5.23 (3.75, 7.30)

Number of symptoms at testingb

 0

Ref

<0.001

Ref

<0.001

Ref

<0.001

0.148

 1–4

0.43 (0.32, 0.57)

 

0.41 (0.24, 0.70)

 

0.37 (0.26, 0.52)

  

 5+

1.88 (1.62, 2.19)

2.20 (1.55, 3.12)

1.49 (1.24, 1.79)

Mobilityd

 No problems

Ref

<0.001

Ref

<0.001

Ref

<0.001

0.024

 Some/a lot of problems

6.81 (5.42, 8.55)

 

8.82 (6.53, 11.91)

 

5.17 (3.65, 7.34)

  

Looking after selfd

 No problems

Ref

<0.001

Ref

<0.001

Ref

<0.001

0.205

 Some/a lot of problems

8.89 (6.92, 11.40)

 

10.58 (7.65, 14.62)

 

7.61 (5.13, 11.28)

  

Doing usual activitiesd

 No problems

Ref

<0.001

Ref

<0.001

Ref

<0.001

< 0.001

 Some/a lot of problems

6.52 (5.57, 7.64)

 

10.31 (8.26, 12.88)

 

4.04 (3.22, 5.08)

  

Having paind

 No problems

Ref

<0.001

Ref

<0.001

Ref

<0.001

<0.001

 Some/a lot of problems

9.84 (8.51, 11.38)

 

14.66 (11.94, 18.00)

 

6.65 (5.40, 8.18)

  

Feeling worried/sadd

 No problems

Ref

<0.001

Ref

<0.001

Ref

<0.001

0.006

 A bit

4.05 (3.56, 4.60)

 

5.11 (4.21, 6.21)

 

3.35 (2.81, 3.99)

  

 Very worried/sad

15.12 (12.20, 18.75)

17.54 (13.17, 23.36)

15.38 (10.90, 21.70)

  1. aP value for interaction between potential predictor and SARS-CoV-2 status (derived in total population)
  2. bWe show (for simplicity) univariable associations for age and number of symptoms modelled as categorical variables. In the final multivariable model, they have been modelled with linear and quadratic terms or a fractional polynomial; see the “Methods” section for details
  3. cFrom the UCLA Loneliness scale. dFrom the EQ-5D-Y scale
Back to article page

BMC Medicine

ISSN: 1741-7015

Contact us