5.3. Childhood single-parent family in relation to criminality in adulthood (III, IV)

Out of the 607 cohort male offenders (at least one registered criminal offence), 212 (34.9 %) were violent and 395 (65.1 %) non-violent criminals. Nearly half of the violent male offenders (n= 98, 46.2 %) had repeated violent acts. Among females, there were 72 offenders, of whom 63 (87.5 %) were non-violent criminals. Of the violent female offenders, four were recidivists. The frequency and percentage distributions of sociodemographic and clinical variables among non-violent, violent and non-criminals are shown in Table 10 for males and in Table 11 for females.

Table 10. Frequency and percentage distributions of some sociodemographic and clinical variablesamong male cohort members according to criminality*.

Variable

Non-violent offenders n (%)Violent offendersNon-criminals n (%)
 All cases n (%)Recidivists n (%) 

Family type

    Single-parent

       All time

9 (2.3)

13 (6.1)

9 (9.2)

58 (1.2)

       At birth

11 (2.8)

15 (7.1)

10 (10.2)

88 (1.8)

       Parental death

35 (8.9)

22 (10.4)

7 (7.1)

296 (5.9)

       Parental divorce

64 (16.2)

36 (17.0)

15 (15.3)

391 (7.8)

    Two-parent

276 (69.9)

126 (59.4)

57 (58.2)

4149 (83.3)

Maternal age

   <20

39 (9.9)

32 (15.1)

17 (17.3)

310 (6.2)

    ≥20

353 (90.1)

180 (84.9)

81 (82.7)

4653 (93.8)

Maternal smoking

   Yes

86 (21.9)

61 (29.8)

31 (32.6)

682 (14)

    No

303 (78.1)

144 (70.2)

64 (67.4)

4196 (86)

Psychiatric hospital diagnosis

   Yes

62 (15.7)

54 (25.5)

33 (33.7)

148 (3.0)

   No

333 (84.3)

158 (74.5)

65 (66.3)

4834 (97.0)

Parental social class

    I, II

99 (25.1)

40 (19.0)

15 (15.5)

1565 (31.4)

   III, IV

276 (69.9)

153 (72.5)

75 (77.3)

2795 (56.1)

    Farmers

20 (5.1)

18 (8.5)

7 (7.2)

619 (12.4)

Perinatal complications

    Yes

24 (6.1)

18 (8.5)

12 (12.2)

353 (7.1)

    No

371 (93.9)

194 (91.5)

86 (87.8)

4629 (92.9)

* Total n on which percentages are based varies due to missing data for some subjects.

Table 11. Frequency and percentage distributions of some sociodemographic and clinical variables among female cohort members according to criminality*

Variable

Non-violent offenders n (%)Violent offendersNon-criminals n (%)
 All cases n (%)Recidivists n (%) 

Family type

    Single-parent

       All time

2 (3.2)

-

-

79 (1.5)

       At birth

6 (9.5)

1 (11.1)

1 (25.0)

103 (2.0)

       Parental death

5 (7.9)

3 (33.3)

2 (50.0)

342 (6.5)

       Parental divorce

15 (23.8)

1 (11.1)

-

475 (9.0)

   Two-parent

35 (55.6)

4 (44.4)

1 (25.0)

4274 (81.1)

Maternal age

    <20

9 (14.3)

-

-

344 (6.6)

    ≥20

54 (85.7)

9 (100.0)

4 (100.0)

4901 (93.4)

Maternal smoking

    Yes

15 (24.6)

3 (33.3)

1 (25.0)

725 (14.1)

   No

46 (75.4)

6 (66.7)

3 (75.0)

4428 (85.9)

Psychiatric hospital diagnosis

   Yes

11 (17.5)

2 (22.2)

2 (50.0)

109 (2.1)

   No

52 (82.5)

7 (77.8)

2 (50.0)

5164 (97.9)

Parental social class

    I, II

11 (17.5)

2 (22.2)

-

1584 (30.1)

    III, IV

49 (77.8)

7 (77.8)

4 (100.0)

3037 (57.6)

    Farmers

3 (4.8)

-

-

648 (12.3)

Perinatal complications

    Yes

7 (11.1)

-

-

411 (7.8)

    No

56 (88.9)

9 (100.0)

4 (100.0)

4862 (92.2)

* Total n on which percentages are based varies due to missing data for some subjects.

5.3.1. Criminality among males

5.3.1.1. Family background and non-violent offences

A single-parent family of type “all time“ (adj. OR 1.6, 95% CI 0.7-3.5) or “at birth“ (adj. OR 1.5, 95% CI 0.7-2.9) was not associated with non-violent offending (III: Table 1). If the parents were divorced during the offspring’s childhood, the male offspring’s risk of committing non-violent crimes was two-fold (adj. OR 2.0, 95% CI 1.5-2.7) compared to males with two-parent family background. Parental death also increased the risk of committing a non-violent crime during early adulthood among males (OR 1.5, 95% CI 1.0-2.3, adjusted for maternal age, maternal smoking during pregnancy, psychiatric hospital diagnosis, parental social class, and perinatal complications).

5.3.1.2. Family background and violent offences

Male subjects who were born and raised in a single-mother family (“all time“) had the highest risk of committing a violent crime during the follow-up time (from 15 to 32 years of age) (III: Table 1). The risk was 5-fold compared to males with a two-parent family background even after controlling for maternal age, maternal smoking during pregnancy, psychiatric hospital diagnosis, parental social class, and perinatal complications (adj. OR 5.2, 95% CI 2.5-10.6). Correspondingly, the probability of violent crimes was 3.6-fold (95% CI 1.8-7.0) if the male subject was born to a single-mother family (“at birth“) and 2.5-fold (95% CI 1.6-3.7) if parents got separated during the follow-up time. If a parent had died during the offspring’s childhood, the risk of violent offending doubled (adj. OR 2.2, 95% CI 1.3-3.6) compared to males with intact, two-parent family backgrounds.

5.3.1.3. Family background and recidivism

The risk of recidivism was highest among males who were born and raised in a single-mother family “all time“, adjusted odds ratio being 7.8 (95% CI 3.1-19.4) (IV: Table 1). Correspondingly, the risk of violent recidivism was 5.0-fold (95% CI 2.1-11.9) if the male subject was born to a single-mother who married later during the offspring’s childhood (“at birth“) and 2.0-fold (95% CI 1.1-3.8) if the parents had a divorce during the offspring’s childhood. Furthermore, parental death was not significantly associated with the risk of violent recidivistic offending among males.

5.3.2. Criminality among females

Females with a single-parent family in childhood had a 2.7-fold risk (95% CI 1.5-4.5) (adjusted for maternal age, maternal smoking during pregnancy, psychiatric diagnosis, parental social class, and perinatal complications) of becoming non-violent criminals over females with two-parent family backgrounds. The risk of violent offending in females with a single-parent family background was 4.4-fold (95% CI 1.1-17.0) compared with that of other females. Due to the small number of cases no subgroup analyses of different family types were made.