Table 3 Association between caffeine metabolic ratios, risk of at least one hospital admission and length of stays in PsyMetab.

From: Dietary caffeine to assess CYP1A2 activity, tailor clozapine doses, and predict treatment response: genetic, epigenetic and clinical analyses

 

Hospital admission

Lengths of stay

Predictors

Odds Ratios

95% CI

p-value

Hazard Ratiosc

95% CI

p-value

(Intercept)

0.06

0.02–0.25

<0.001

   

Age (10 years)

1.56

1.42–1.72

<0.001

0.87

0.84–0.90

<0.001

Sex (Male)

0.76

0.57–1.01

0.059

1.11

0.99–1.25

0.13

Psychiatric diagnosisa

      

 Mood disorders

1.54

1.06–2.26

0.02

1.00

0.83–1.21

0.99

 Psychotic disorders

5.88

3.86–9.05

<0.001

0.87

0.72–1.06

0.28

Smoking

0.99

0.74–1.34

0.97

1.14

1.01–1.28

0.05

Caffeine metabolic ratiosb

1.26

1.09–1.46

0.002

0.89

0.84–0.95

<0.001

rs2472297 (TT or CT; ref: CC)

0.85

0.61–1.19

0.35

0.98

0.86–1.12

0.79

rs4410790 (TT or CT; ref: CC)

1.34

0.99–1.82

0.057

0.99

0.88–1.11

0.88

rs56113850 (TT or CT; ref: CC)

1.21

0.36–4.01

0.75

0.82

0.44–1.54

0.38

Clozapine or olanzapine use

1.59

1.05–2.46

0.03

0.88

0.77–1.01

0.11

Patients

1019

664

Observations

1019

1349

  1. Significant p-values in bold. Psychiatric diagnoses were defined as follow: psychotic disorders [F20-F25 and F28-F29], Mood disorders [F30-F31 and F32-F33], and other diagnoses [F00-F19, F34- F99].
  2. aCompared to other diagnoses.
  3. bCalculated by dividing paraxanthine by caffeine plasma levels and multiplying the ratio by 1.07769 (Molecular mass ratio).
  4. cHazard ratios > 1 indicated the higher likelihood of a short length of stay (i.e., length of stay below the median of length of stays of 42 days), whereas Hazard ratio < 1 indicated the risk of a prolonged length of stay (i.e., length of stay equal to or higher than the median of length of stays of 42 days).
  5. CI Confidence Interval, ref reference.