TABLE 5.
Unit-weighted scores |
Unit-weighted score differencesa |
Latent mean differencesb |
||||||
Category | Subgroup | N | Mean | SD | SES | P-value | SES | P-value |
PHQ-9 | ||||||||
Total | 1207 | 3.71 | 3.72 | |||||
Gender | Female | 553 | 4.66 | 3.93 | ||||
Male | 654 | 2.91 | 3.34 | −0.417 | <0.001 | −0.429 | <0.001 | |
Age | 30–40 | 326 | 3.24 | 3.27 | ||||
41–50 | 493 | 3.66 | 3.78 | 0.076 | 0.203 | 0.094 | 0.228 | |
51–60 | 388 | 4.16 | 3.97 | 0.152 | 0.016 | 0.194 | 0.001 | |
Education | Primary | 305 | 4.83 | 4.34 | ||||
Secondary | 619 | 3.72 | 3.67 | −0.195 | 0.001 | −0.259 | <0.001 | |
Higher | 283 | 2.45 | 2.56 | −0.446 | <0.001 | −0.602 | <0.001 | |
Status | At risk | 1006 | 3.77 | 3.77 | ||||
T2D | 201 | 3.4 | 3.47 | −0.046 | 0.477 | −0.085 | 0.508 | |
Time | T0 | 1006 | 3.77 | 3.77 | ||||
T1 | 982 | 3.09 | 3.49 | −0.168 | <0.001 | −0.151 | <0.001 | |
T2 | 963 | 2.68 | 3.19 | −0.255 | <0.001 | −0.262 | <0.001 | |
GAD-7 | ||||||||
Total | 1205 | 3.16 | 3.33 | |||||
Gender | Female | 554 | 3.9 | 3.72 | ||||
Male | 651 | 2.52 | 2.8 | −0.299 | <0.001 | −0.603 | <0.001 | |
Age | 30–40 | 324 | 3.09 | 3.19 | ||||
41–50 | 495 | 3.09 | 3.26 | −0.039 | 0.471 | 0.042 | 0.583 | |
51–60 | 386 | 3.3 | 3.53 | 0.002 | 0.973 | 0.130 | 0.101 | |
Education | Primary | 305 | 3.91 | 3.81 | ||||
Secondary | 617 | 3.1 | 3.2 | −0.144 | 0.007 | −0.292 | <0.001 | |
Higher | 283 | 2.47 | 2.86 | −0.328 | <0.001 | −0.541 | <0.001 | |
Status | At risk | 1003 | 3.21 | 3.37 | ||||
T2D | 202 | 2.88 | 3.11 | −0.065 | 0.265 | −0.138 | 0.181 | |
Time | T0 | 1003 | 3.21 | 3.37 | ||||
T1 | 981 | 2.62 | 2.94 | −0.132 | <0.001 | −0.180 | <0.001 | |
T2 | 963 | 2.42 | 2.81 | −0.172 | <0.001 | −0.256 | <0.001 |
aBased on robust regression with MM-estimation.
bBased on multigroup structural equation modeling using a second order model for the PHQ-9 to allow for comparison of the general factor latent mean. SD, standard deviation; SES, standardized effect size; T2D, type two diabetes. Interpretation of the SES corresponds to Cohen’s d (conventionally: d = 0.20, 0.50, and 0.80 for small, medium, and large effects, respectively). Empty rows indicate the reference category.