Table 2

Meta-regression analyses of sources of heterogeneity in the prevalence of psychosis and major depression in prisoners

PsychosisMajor depression
Variable and study characteristicax03B2;s.e.(x03B2;)Px03B2;s.e.(x03B2;)P
Gender of inmates: male v. female0.00160.0630.8000.03230.02220.151
Mean age of inmates (continuous)–0.00090.00090.3340.00110.00330.735
Year of study (continuous)<0.00010.00030.8890.00150.00130.340
Country
 Low/middle v. high income0.02040.00950.0350.11570.03180.001
 USA v. rest of the world0.00070.00600.902–0.00430.02410.859
 Within the USA, over time–0.00060.00050.2410.00380.00130.008
Prisoner status: sentenced prisoners v. detainees0.00250.00380.5040.01360.01590.396
On reception: first 2 weeks of reception v. rest–0.00160.00580.778–0.01310.02410.588
Participation rate
 Continuous–0.01840.03330.585–0.07780.09980.441
 Low (⩽80%) v. high (>80%)–2.6373.09640.400–0.27861.14580.809
Sample size
 Continuous<0.00010.00000.337<0.00010.00000.577
 ⩽500 v. >5000.00880.00620.156–0.01470.02750.594
Interviewer: psychiatrist v. other–0.00150.00530.7840.01990.02310.391
Diagnostic criteria: ICD v. DSM–0.00210.00550.706–0.05900.0250.021
  • Significant associations (P<0.05) are in bold.

  • a. For comparisons the reference category is given first.