Table 2 Table of GEE models for outcome of MDD symptoms and diagnoses.

From: Loneliness associates strongly with anxiety and depression during the COVID pandemic, especially in men and younger adults

 

MDD symptoms

MDD diagnoses

Main model

Model including PRSa

Main model

Model including PRSa

Terms

IRR [95% CI]

IRR [95% CI]

OR [95% CI]

OR [95% CI]

Time

0.992 [0.991, 0.994]*

0.990 [0.987, 0.993]*

1.016 [1.009, 1.022]*

1.020 [1.007, 1.034]

Time2

1.000 [1.000, 1.000]*

1.000 [1.000, 1.000]*

1.000 [1.000, 1.000]*

1.000 [0.999, 1.000]

Loneliness

1.275 [1.251, 1.300]*

1.271 [1.225, 1.320]*

1.559 [1.469, 1.654]*

1.539 [1.365, 1.734]*

Lone × time

1.001 [1.000, 1.001]*

1.001 [1.000, 1.001]

1.001 [1.000, 1.001]

1.001 [1.000, 1.002]

Lone × age

0.999 [0.999, 0.999]*

0.999 [0.999, 1.000]

1.000 [0.999, 1.002]

1.002 [0.999, 1.004]

Lone × sex

0.971 [0.961, 0.981]*

0.964 [0.945, 0.985]*

0.967 [0.936, 0.998]

0.920 [0.859, 0.986]

Lone × MDD PRS

 

0.994 [0.985, 1.002]

 

0.978 [0.950, 1.006]

Age

0.982 [0.981, 0.983]*

0.983 [0.981, 0.985]*

0.967 [0.964, 0.971]*

0.965 [0.958, 0.971]*

Female sex

1.397 [1.357, 1.437]*

1.367 [1.288, 1.450]*

1.228 [1.112, 1.356]*

1.311 [1.055, 1.631]

MDD PRS

 

1.116 [1.087, 1.145]*

 

1.265 [1.151, 1.391]*

Constant

0.980 [0.923, 1.040]

0.894 [0.799, 1.001]

0.050 [0.041, 0.061]*

0.043 [0.030, 0.063]*

N

75,279

19,128

75,279

19,128

CIC

16.222

32.889

15.8

32.793

  1. Displayed are incidence rate ratios (IRR) and odds ratios (OR) with 95% confidence intervals for the different terms in the GEE models. Asterisks indicate statistical significance at Bonferroni-corrected α = 0.00125. All models include a linear and squared time term, a main loneliness effect and a loneliness-time interaction. Furthermore, terms for age and female sex are included both as main effect and a loneliness interaction term. The models in genotyped samples also include an MDD polygenic risk score both as main effect and as a loneliness interaction. N refers to the number of individuals included in the dataset. The correlated information criterion (CIC) is a measure of fit for GEE models.
  2. aIn sample with genotype data available. Models include the first 10 principle components of the genotype data as covariates as well (not shown).
Back to article page