Factors associated with internalizing or somatic symptoms in a cross-sectional study of school children in grades 1-10

Children's perceived health status influences their daily life [1, 2], and childhood health is also a powerful predictor for health in adulthood [3, 4]. Health complaints are typically classified as either emotional or somatic, and a combination of these types of symptoms is not uncommon [5‐10].

Anxiety and depression are the most common emotional problems, and appear to be more prevalent among girls, with fairly high co-morbidity (20-50%) [11]. Anxiety tends to predate depression [6, 9], and the prevalence may range from 6% to 18% in childhood and adolescence [11]. Depressive disorders are rare among young children, but in adolescence the prevalence may be as high as 8% [11]. The results of long term follow-up studies suggest that early emotional symptoms may predict higher risk of mental and physical disease in middle age [12‐14].

Headache and stomach pain are the most prevalent physical complaints at a young age [15]. Before elementary school, children rarely complain about headache [16], but the prevalence increases with age [10, 17, 18]. Around puberty, about 15% may report frequent or severe headache, and more than half of the students in high school may report less frequent episodes of headache [17]. Before puberty, the prevalence of reported headache seems to be higher in boys than girls, but after puberty, the prevalence appears to be higher among girls [17, 18].

Stomach pain appears to be more frequent among younger than older children [16, 19, 20]. Recurrence of abdominal pain may range from 10-45% [21], and in adolescence (11-15 years), the total prevalence of self-reported episodes of stomach pain is around 50%, and the estimates are higher for girls than boys [20, 22]. Perceived abdominal pain in childhood has been associated with higher risk of both physical and mental disorders later in life [23, 24].

In school, both circumstances in class and during recess may be important for the children's health and wellbeing. Learning disabilities, low academic achievement or emotional distress may be associated with poorer health [22, 25‐28]. Victimization caused by bullying, as well as perceived loneliness, have also been associated with adverse health effects, both in the short and long term [29‐31]. Further, experiencing caring teachers and belonging to school have been related to good health and wellbeing [32], and negatively associated with emotional distress and risky behaviour [33, 34]. There is also evidence to suggest that connectedness to school may be associated with better health in the long term and less risky behaviour [35‐37].

In the present study of more than 400 school children, we collected information on self-reported sadness and anxiety, and headache and stomach ache. The aim was to assess whether factors assumed to influence health status, either negatively or positively, were associated with the prevalence of the four symptoms.

Among 419 participating children (230 boys and 189 girls), gender was evenly distributed by school grade [38]. Table 1 shows children's scores for the dependent variables; sadness, anxiety, stomach ache and headache, as well as for each independent, potentially explanatory factor. Most of the children (between 67% and 83%) reported never or seldom to have experienced any of the four symptoms, whereas about one in four children had experienced one or more of the symptoms now and then or more often. The score distribution for the independent variables was similar to the distribution of the outcomes, with the majority of children reporting the two best scores.

We assessed the association of each independent variable with the respective scores for sadness, anxiety, stomach ache and headache. The left part of Table 2, 3, 4, and 5 show the association of each independent variable, with adjustment for gender and grade. In the right part of the tables, the associations are also adjusted for the other variables listed in the table.

In this cross-sectional study of self-reported internalizing and somatic symptoms among more than 400 school children, we found that perceived loneliness was strongly associated with the prevalence of sadness, anxiety and headache, also after adjustment for a number of potentially confounding factors. In separate analyses of boys and girls, loneliness in boys was strongly associated with sadness, whereas in girls, the association of loneliness was equally strong for sadness, anxiety and headache.

The associations of loneliness were robust, and did not substantially change from the crude (only adjusting for grade and gender) to the multivariable analysis. The results suggest that loneliness may be particularly important among girls, since loneliness was the most important correlate to high scores for three of the four symptoms.

The study was conducted in public schools in rural communities, ranging from inland to coastal environments. The population base and the very high attendance are strengths of the study, but it is a weakness that children from urban settings were not included. In the data collection, younger children were interviewed by school nurses, whereas older children completed the questionnaire themselves. Although the nurses were trained for this task, the possibility that the different procedures could have influenced the responders and introduced systematic differences in results between younger and older children can not be excluded.

Also, the cross-sectional design is a limitation of this study. That the children simultaneously reported exposures and outcomes may lead to inter-related responses to the questions, and could have caused stronger associations between explanatory factors and health outcomes. Thus, collecting outcomes at a later stage could have yielded different results. Therefore, the findings should be interpreted with caution, since cross sectional designs limit the possibility to study causal effects.

The internalizing and somatic symptoms that we used as outcome measures in this study are common, and there is evidence suggesting that self-reports of emotional and somatic symptoms are reasonably reliable in studies of health in adolescence [42]. Internalizing and somatic symptoms may infer with children's daily living and cause absence from school [1]. Further, previous studies of internalizing or somatic symptoms in childhood and adolescence have shown an increased risk of anxiety disorders, depression, and somatic illness later in life [3, 4, 9, 12, 14, 23, 24].

In the initial analyses (only adjusting for gender and grade) among factors that were assumed to promote health, children's satisfaction with academic work and the help they receive from teachers were associated with a relatively lower prevalence of symptoms. After mutual adjustment for other variables, only the negative association of help from teachers with stomach ache in girls remained significant. Previously, it has been suggested that academic satisfaction may be beneficial for children's health [43], and that support from teachers may provide protection against poor health [35, 44].

Each factor that was assumed to be adversely related to health was associated with higher scores for each of the four symtoms in the crude analyses, but after mutual adjustment for other potentially explanatory variables, most of the initial associations were fully attenuated. In other studies, multivariable adjustment also attenuated the estimates, but to different degrees [42, 45‐47]. Victimization caused by bullying is an example of a factor that has shown robust associations, also in multivariable analyses.

In this study, loneliness was the only factor that retained the strong relation to poorer health after adjustment for other potentially confounding factors. We cannot rule out the possibility that factors that we failed to include in the study, at least in part, may explain the associations of loneliness. Thus, it has been suggested that close friendship and peer acceptance could modify effects related to loneliness [48‐50]. On the other hand, it may be equally plausible that the variable loneliness captures something that in itself is strongly associated with the internalizing and somatic symptoms that we have studied. Sadness may be a key emotion for both depression [6, 51, 52] and loneliness [48], but the link of loneliness to the physical complaints, headache and stomach pain, may not be easily explained, unless these complaints represent somatic expressions of underlying emotional distress [5, 7, 53].

Only a few studies have assessed the association of perceived loneliness with health problems in childhood and adolescence, and to our knowledge, no previous study has assessed loneliness in relation to headache or stomach pain. Nonetheless, the strong associations that we found for loneliness and emotional distress are in line with previous findings. In cross-sectional studies, it has been suggested that loneliness is associated with both anxiety [54, 55] and depression [30, 56], and that persistent loneliness may contribute to later emotional disorders [56]. From a recent prospective study that followed children from childhood to adolescence, it was reported that measures of loneliness at the age of 5 and 9 years could predict depressive symptoms at 13 years of age [57].

Few studies have compared internalizing or somatic symptoms between girls and boys in relation to loneliness, and there are no consistent gender differences [56]. We found, however, a strong association of loneliness with anxiety and headache among adolescent girls, but not in boys, whereas for sadness, there was a clear association of loneliness for both genders.

In this population study of children between 7 and 16 years of age, perceived loneliness appears to be of special importance in relation to internalizing and somatic symptoms, and for girls, perceived loneliness may be particularly important in relation to emotional distress (sadness and anxiety) and physical complaints (headache). Longitudinal studies that measure the impact of factors that are associated with perceived loneliness and their relation with subsequent health problems are recommended.

Emotional and somatic symptoms are common in childhood and adolescence. Teachers, school nurses, clinicians, and others need to be aware of the strong relation between loneliness and ill health, and daily routines should be established to reduce loneliness among school children. It is possible that a caring attention from teachers and school nurses combined with strategic planning of activities and peer collaboration may reduce loneliness among the children.