In Germany, studies focusing on the course of MHP from childhood into adolescence are insufficient. For this reason, this study within the Future Family project tried to examine how frequent and stable MHP are from kindergarten age to adolescence over the course of 10 years. We expected that about 18% of children are affected by MHP and that MHP have a persistence rate of around 50%. Every tenth child should show a chronic mentally ill course.
Prevalence of mental health problems (MHP) in our sample in childhood and adolescence
Approximately 24% of the sample showed borderline clinical or clinically relevant MHP. Internalizing MHP seem to occur slightly more frequently than Externalizing MHP. These rates are similar to the results of the KiGGS-study, which found 20% borderline and clinically relevant MHP [11]. If only the clinically relevant rates are considered, a prevalence rate of about 15% was found, which is slightly higher than the worldwide (13%) [1] and slightly lower than the prevalence rate reported in Germany (17–18%) [2, 3]. Furthermore, when observing the course of MHP, it is apparent that the prevalence rates decrease overall with increasing age in childhood, while a large rise, especially in Internalizing MHP, was found in puberty. Thus, the onset of adolescence seems to be a particularly vulnerable time in child development.
Non-significant differences in prevalence rates between boys and girls were found. Nevertheless, there was a great increase in Internalizing MHP for boys during puberty. The assumption that boys in Germany show higher prevalence rates than girls until puberty, therefore, could not be confirmed [3, 5, 7, 9, 12].
Persistence of mental health problems in childhood and adolescence
Regarding the first method using course-types, we found similar results for all three analyzed categories. Approximately 70% of the CA never showed any or only once MHP, whereas 14–15% were chronic mentally ill. While we found comparable frequencies to the KiGGS-sample (74%) [5] for the stable healthy group, our sample showed slightly more chronic mentally ill courses than previously assumed by Esser et al. (10%) [4]. Furthermore, the same number of children (3–4%) experienced a positive or negative course. About 8–10% of our sample showed transient courses meaning that their MHP were only clinically relevant at some assessment points. This could provide evidence that a greater proportion of CA suffer from recurrent MHP. Referring to sex-specific aspects, significant different distributions were only found for Externalizing MHP. Boys showed significantly more often a negative or positive course, whereby both course-types were represented equally frequently.
Our second approach was methodologically oriented to Esser et al. [4] and their transition probabilities. In the following results, a distinction between short- (from one assessment point to the next one) and long-term (from FU2 to FU10 = from childhood into adolescence) persistence should be made. Regarding the short-term persistence, similar percentages of mostly around 60–70% were found for the Total Score as well as Internalizing and Externalizing MHP. Our results differ from the BELLA-study [5], which reported persistence values from 30 to 50%. The reason for this can be found in the method. In the BELLA-study CA from 7 to 17 years formed the baseline sample, whereas in this study the course of kindergarten children from 3 to 5 years was analyzed. Accordingly, in this sample the course of MHP is considered in a sample of children with similar age. Another explanation for these high persistence values is the combination of borderline clinical and clinically relevant values leading to higher prevalence rates. Furthermore, hints for a decreasing tendency with age was found, which was also visible in the BELLA-study. The short-term persistence should be rated as high, taking into account the short time intervals of only 1 year between most assessment points.
Long-term persistence rates from childhood to adolescence were highest for Internalizing MHP (59%), while the Total Score (54%) and Externalizing MHP (51%) showed slightly lower rates. Overall, these results are in line with the findings from Esser and colleagues [4] as well as Ihle and Esser [9], reporting persistence rates of more than 50%. Contrary to our assumption, Internalizing MHP were more stable than Externalizing MHP. When looking at the sex-specific differences, particularly clinically relevant values in childhood reported by boys showed high persistence rates up to puberty for all three categories. While the long-term persistence rates from girls were around 50%, boys MHP were more stable, both in the Total Score (59%) and Internalizing MHP (68%).
For the prediction of chronic courses, MHP of the mother in particular proved to be a decisive predictor, whereas child’s intelligence, mother’s education, dysfunctional parenting, and sex showed no significant influence in these models. The finding that MHP of the mother contribute to the continuity of pathology in their children is consistent with the results of O’Connor and colleagues [17] identifying depression in the primary caregiver as the only significant familial variable predicting the persistence. On the other hand, the sex and cognitive abilities of the child as well as socioeconomic aspects showed significant results in their study, which is in contrast to our results. To prevent chronic mentally ill courses in children, the mental health of the mother could be a significant starting point for interventions.
Strengths and limitations
On the positive side, data were collected from kindergarten children up to adolescence, with three different categories of MHP being analyzed simultaneously. Furthermore, ideas for the analysis of persistence in the form of number chains (0 0 0 1 1) and the distinction between short- and long-term persistence were presented.
On the other hand, several limitations should be discussed. First, only the mother’s perspective was analyzed. Second, since the sample is mainly from Brunswick and comes from the higher social class, which could lead to lower MHP [3], the generalizability of the results is limited. Third, for the CBCL1½—5 no German norms are available which is why the American norms were used for the pre-assessment point. Due to the more liberal American norms at pre, the reported prevalence rates were significantly lower and an increase in MHP at FU2, when German norms were used, is noticeable. This inevitably had a direct impact on the analysis of the number sequences and the persistence rates. Additionally, a huge proportion (54%) of the parents in our sample participated in the Triple P [20], a parenting program reducing dysfunctional parenting and preventing MHP in CA leading to altered and potentially lower prevalence and persistence rates. Taking these aspects into account, it can be assumed that the number of clinically relevant MHP should be even higher than reported in this study. Finally, the co-occurrence of Internalizing and Externalizing MHP was not further analyzed, while it could play an important role in child development and the persistence of MHP [36].