EPIDGS32 BASIC STATISTICS FOR MEDICAL SCIENCES

Assessment 1 – data analysis

Table 1. Descriptive characteristics of the sample

Variable Men Women Total

(N=1620) (N=1861) (N=3481)
Psychological distress Yes 264 (16.3%) 578 (31.1%) 842 (24.2%)
No 1356 (83.7%) 1283 (68.9%) 2639 (75.8%)
GHQ-12 Mean (SD) 8.6 (5.0) 10.9 (5.8) 9.8 (5.5)
Maternal age Up to 20 150 (9.3%) 118 (6.3%) 268 (7.7%)
21-25 413 (25.5%) 529 (28.4%) 942 (27.1%)
26-30 560 (34.6%) 758 (40.7%) 1318 (37.9%)
31-35 371 (22.9%) 300 (16.1%) 671 (19.3%)
>35 126 (7.8%) 156 (8.4%) 282 (8.1%)
Household income Low 1233 (76.1%) 1442 (77.5%) 2675 (76.9%)
High 387 (23.9%) 419 (22.5%) 806 (23.2%)
Marital status Married 1322 (81.6%) 1463 (78.6%) 2785 (80.0%)
Cohabiting 116 (7.2%) 167 (9.0%) 283 (8.1%)
Single 182 (11.2%) 231 (12.4%) 413 (11.9%)
Maternal employment Yes 800 (49.4%) 1010 (54.3%) 1810 (52.0%)
No 820 (50.6%) 851 (45.7%) 1671 (48.0%)
Maternal smoking Yes 523 (32.3%) 625 (33.6%) 1148 (33.0%)
No 1097 (67.7%) 1236 (66.4%) 2333 (67.0%)
Maternal GHQ Distressed 660 (40.7%) 715 (38.4%) 1375 (39.5%)
Not distressed 949 (58.6%) 1085 (58.3%) 2034 (58.4%)
Missing 11 (0.7%) 61 (3.3%) 72 (2.1%)
Table 2. The relationship between GHQ-20/psychological distress and other variables
Variable Category N Continuous GHQ-20 P* Binary distress Chi2 (d.f.)/Chi2 P
Mean (SD) N (%) trend**
Household income Low 2675 9.69 5.63 0.003 620 23.2 6.4 (1) 0.01
High 806 10.36 5.23 222 27.5
Gender Men 1620 8.59 4.99 <0.001 264 16.3 102.9 <0.001
Women 1861 10.94 5.77 578 31.1
Maternal age Up to 20 268 9.79 6.16 0.001 62 23.1 19.3 (4) 0.001
9.1 (1) 0.003
21-25 942 9.78 5.80 221 23.5
26-30 1318 9.47 5.33 284 21.6
31-35 671 10.29 5.33 184 27.4
>35 282 10.80 5.44 91 32.3
Marital status Married 2785 9.76 5.38 0.15 651 23.4 5.0 (2) 0.08
Cohabiting 283 10.08 6.02 77 27.2
Single 413 10.29 6.24 114 27.6
Maternal employment Yes 1810 9.86 5.36 0.86 450 24.9 0.93 (1) 0.33
No 1671 9.83 5.74 392 23.5
Maternal smoking Yes 1148 10.19 5.76 0.01 299 26.1 3.22 (1) 0.07
No 2333 9.68 5.43 543 23.3
Maternal GHQ Distressed 1375 10.24 5.60 <0.001 434 21.3 21.9 (2) <0.001
Not distressed 2034 9.53 5.49 386 28.1
Missing 72 11.24 5.51 22 30.6
* unpaired t-test or one-way ANOVA ** when appropriate
Table 3. Association between household income and risk factors associated with psychological
distress
High income
Variable Category N % Chi2 (d.f.) P
Age Up to 20 18 6.7 445.7(4) <0.001
21-25 75 8.0 383.9(1) <0.001
26-30 275 20.9
31-35 312 46.5
>35 126 44.7
Gender Males 387 23.9 0.91 0.34
Females 419 22.5
Maternal marital status Married 701 25.2 73.5 (2) <0.001
Cohabiting 78 27.6
Single 27 6.5
Maternal GHQ Distressed 492 24.2 <0.001*
Not distressed 314 22.8
Missing 0 0.0
Maternal smoking Yes 160 13.9 81.8(1) <0.001
No 646 27.7
*Fisher’s exact test
Table 4. OR of psychological distress by household income, adjusted for confounders and
stratified by effect modifiers
Level of adjustment OR (95% CI) p-value p-value
(hh_income) homogeneity
Crude (unadjusted) 1.26 1.05-1.51 0.01
Age 1.15 0.95-1.40 0.14 0.17
Maternal smoking 1.30 1.09-1.56 0.004 0.20
Maternal GHQ 1.29 1.07-1.54 0.006 0.69
Gender* 1.29 1.07-1.54 0.006 0.23
Maternal employment* <0.001
Not employed 1.84 1.39-2.43 <0.001
Employed 0.96 0.76-1.22 0.74
Maternal marital status 1.30 1.08-1.55 0.005 0.62
Adjusted for age, maternal
smoking, maternal marital
status
Not employed 1.39 1.00-1.93 0.05
Employed 0.88 0.68-1.15 0.34
Adjusted for age, maternal
smoking, maternal marital
status, maternal GHQ, gender
Not employed 1.74 1.25-2.41 0.001
Employed 1.00 0.78-1.30 0.97
*these might be effect modifiers although they are unlikely confounders
There is strong interaction between maternal employment and household income. Among those whose
mothers did work at age 4, household income does not play any role while among those whose mothers
did not work, their children were more likely to suffer from psychological distress at age16. The effect
presented in the table depends on variables used as confounding variables; two possible adjusted results
shown above.