Table 2

HRs for the risk of Crohn’s disease or UC associated with depression in unadjusted and adjusted models

Crohn’s disease
HR (95% CI)
UC
HR (95% CI)
Unadjusted model
 Depression1.67 (1.44 to 1.93)1.41 (1.29 to 1.55)
Multivariable adjusted model
 Depression2.11 (1.65 to 2.70)2.23 (1.92 to 2.60)
 Age at start of follow-up
  Under 50 years1.00 (ref)1.00 (ref)
  Over 50 years0.66 (0.56 to 0.78)1.15 (1.05 to 1.25)
 Sex
  Female1.00 (ref)1.00 (ref)
  Male0.71 (0.61 to 0.82)0.94 (0.87 to 1.02)
 Socioeconomic status*
  11.00 (ref)1.00 (ref)
  21.19 (0.97 to 1.45)1.06 (0.95 to 1.18)
  31.23 (1.00 to 1.50)1.00 (0.89 to 1.13)
  41.03 (0.82 to 1.28)0.96 (0.85 to 1.09)
  50.99 (0.77 to 1.27)0.99 (0.86 to 1.14)
 Charlson Comorbidity Index
  01.00 (ref)1.00 (ref)
  >11.33 (1.12 to 1.58)1.19 (1.08 to 1.31)
 Smoking
  Never smoker1.00 (ref)1.00 (ref)
  Ex-smoker1.22 (0.97 to 1.54)1.48 (1.32 to 1.66)
  Current1.64 (1.41 to 1.92)1.09 (0.99 to 1.20)
 Obesity
  Not obese1.00 (ref)1.00 (ref)
  Obese0.89 (0.73 to 1.10)0.82 (0.73 to 0.93)
 Anxiety
  No anxiety1.00 (ref)1.00 (ref)
  Anxiety diagnosis0.72 (0.54 to 0.97)1.03 (0.87 to 1.21)
 Atypical antidepressant
  Referent1.00 (ref)1.00 (ref)
  Medication use1.28 (0.77 to 2.13)0.83 (0.56 to 1.24)
 Mirtazapine
  Referent1.00 (ref)1.00 (ref)
  Medication use0.20 (0.03 to 1.41)0.34 (0.15 to 0.77)
 Monoamine oxidase inhibitor
  Referent1.00 (ref)1.00 (ref)
  Medication useNA 2.77 (0.69 to 11.11)
 Serotonin norepinephrine reuptake inhibitor
  Referent1.00 (ref)1.00 (ref)
  Medication use0.81 (0.36 to 1.81)0.46 (0.25 to 0.83)
 Selective serotonin reuptake inhibitor
  Referent1.00 (ref)1.00 (ref)
  Medication use0.63 (0.50 to 0.78)0.48 (0.42 to 0.55)
 Serotonin modulator
  Referent1.00 (ref)1.00 (ref)
  Medication use0.59 (0.19 to 1.85)0.46 (0.23 to 0.92)
 Tricyclic antidepressant
  Referent1.00 (ref)1.00 (ref)
  Medication use0.77 (0.61 to 0.97)0.59 (0.51 to 0.68)
  • *Socioeconomic status based on the Townsend deprivation quintiles (baseline: least deprivation).

  • †The HR was not able to be estimated due to the small sample sizes of monoamine oxidase inhibitor (MAOI) users.