One mental health problem influencing the risk for another, within individuals and between siblings. Commentary on Allegrini et al. (2022)

Co‐occurrence of mental health symptoms, either at the same time or across development, is a major area of research. It is well known that comorbidity is associated with poorer outcomes and continuation of childhood symptoms into adulthood (Costello & Maughan, 2015). Better insight into mechanisms underlying the development of comorbidity could provide leads for interventions to prevent the continuation of mental disorders. So far, twin studies have mainly focused on genetic and environmental factors influencing the continuation of symptoms or the development of comorbid symptoms over time in childhood (Kuja‐ Halkola et al., 2015; Wertz et al., 2015). Allegrini et al. (2022) have taken a rather novel approach in exploring mechanisms underlying co‐occurrence of symptoms over time. Focusing on social, internalising, attention deficit hyperactivity disorder (ADHD) and externalising symptoms, they not only included between person processes in their longitudinal model, like in the before mentioned studies, but also within person processes to explain the co‐occurrence of symptoms. In between person processes comprise of factors that influence two or more psychopathologies, whereas in within person processes one type of psychopathology influences the other. A genetically informed study applying Mendelian Randomisation analyses suggested that childhood maltreatment is causally associated with depression (unidirectional) as well as with (ADHD) and schizophrenia (bidirectional) (Warrier et al., 2021), that is, childhood maltreatment could be part of the between person effects explaining the comorbidity between these disorders. An example of a within person process is that someone with social anxiety doesn't have friends and then gets depressed because of that. Knowledge on which of these two processes explains the co‐occurrence can aid in choosing the right treatment strategy. While in the first example all disorders need to be addressed in treatment, in the second example a successful treatment of the social anxiety could be sufficient as it will lead to a decrease of the depressive symptoms. Allegrini et al. tested the between and within person processes in a random intercept cross‐lagged panel model (RI‐CLPM) using data collected at age 7, 9/10 and 12 in the Twin Early Development Study (TEDS) (N varies between 5828 and 7758) and the Netherlands Twin Register (NTR) (N varies between 8748 and 12,710). In addition, they extended their model to family data allowing to explore the existence of reciprocal directional influences between siblings (within family effects) besides between family effects, such as genetic effects, to explain resemblance in psychopathology between siblings. The sibling analyses are presented as supplementary analyses, I think, because they are not mentioned in the preregistered protocol at the open science forum, which shows the scientific rigour of the authors. The first analyses where they tested the RI‐CLPM using data from one individual per family showed that the majority of the variance in social, internalizing, ADHD and externalising problems was explained by between person processes, ranging from 42% to 57% in both cohorts. In addition, significant within person effects were detected, explaining between 4% and 11% of the variance. In this commentary, I will only focus on the within person processes that were found in both cohorts as these results seem more robust. These were effects from social problems to internalising problems and from externalising problems to attention and internalising problems for the time lag age 7–9/10 and from internalising problems to social problems for the time lag age 9/10 to age 12. The authors emphasize that these directed influences do not need to be causal influences as there still may be confounding factors at play. They for example, mention that developmental genetic changes, that is, new genetic effects coming into play at a later age, probably would be pushed into the within person processes. If in future analyses, the

Co-occurrence of mental health symptoms, either at the same time or across development, is a major area of research. It is well known that comorbidity is associated with poorer outcomes and continuation of childhood symptoms into adulthood (Costello & Maughan, 2015).
Better insight into mechanisms underlying the development of comorbidity could provide leads for interventions to prevent the continuation of mental disorders.
So far, twin studies have mainly focused on genetic and environmental factors influencing the continuation of symptoms or the development of comorbid symptoms over time in childhood (Kuja-Halkola et al., 2015;Wertz et al., 2015). Allegrini et al. (2022) have taken a rather novel approach in exploring mechanisms underlying co-occurrence of symptoms over time. Focusing on social, internalising, attention deficit hyperactivity disorder (ADHD) and externalising symptoms, they not only included between person processes in their longitudinal model, like in the before mentioned studies, but also within person processes to explain the co-occurrence of symptoms. In between person processes comprise of factors that influence two or more psychopathologies, whereas in within person processes one type of psychopathology influences the other. A genetically informed study applying Mendelian Randomisation analyses suggested that childhood maltreatment is causally associated with depression (unidirectional) as well as with (ADHD) and schizophrenia (bidirectional) (Warrier et al., 2021), that is, childhood maltreatment could be part of the between person effects explaining the comorbidity between these disorders. An example of a within person process is that someone with social anxiety doesn't have friends and then gets depressed because of that. Knowledge on which of these two processes explains the co-occurrence can aid in choosing the right treatment strategy. While in the first example all disorders need to be addressed in treatment, in the second example a successful treatment of the social anxiety could be sufficient as it will lead to a decrease of the depressive symptoms.
Allegrini et al. tested the between and within person processes in a random intercept cross-lagged panel model (RI-CLPM) using data collected at age 7, 9/10 and 12 in the Twin Early Development Study

(TEDS) (N varies between 5828 and 7758) and the Netherlands Twin
Register (NTR) (N varies between 8748 and 12,710). In addition, they extended their model to family data allowing to explore the existence of reciprocal directional influences between siblings (within family effects) besides between family effects, such as genetic effects, to explain resemblance in psychopathology between siblings. The sibling analyses are presented as supplementary analyses, I think, because they are not mentioned in the preregistered protocol at the open science forum, which shows the scientific rigour of the authors.
The first analyses where they tested the RI-CLPM using data from one individual per family showed that the majority of the variance in social, internalizing, ADHD and externalising problems was explained by between person processes, ranging from 42% to 57% in both cohorts. In addition, significant within person effects were detected, explaining between 4% and 11% of the variance. In this commentary, I will only focus on the within person processes that were found in both cohorts as these results seem more robust.
These were effects from social problems to internalising problems and from externalising problems to attention and internalising problems for the time lag age 7-9/10 and from internalising problems to social problems for the time lag age 9/10 to age 12. The authors emphasize that these directed influences do not need to be causal influences as there still may be confounding factors at play.
They for example, mention that developmental genetic changes, that is, new genetic effects coming into play at a later age, probably would be pushed into the within person processes. If in future analyses, the This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.