Impact of the first COVID-19 lockdown on the relationship with parents and peers in a cohort of adolescents with somatic symptom disorder

Italy is one of the countries most affected by the Coronavirus Disease 2019 (COVID-19) pandemic [1]. Due to the exponential spread of infected people and proportional deaths, the Italian government imposed a national lockdown from March 9 to May 4, 2020. As part of the lockdown process, all schools were closed, extracurricular and sports activities were suspended, and people movement from home was allowed only for emergencies. Therefore, children and adolescents were subjected to a strict ban on leaving their homes without justified reasons.

The reported mental health consequences of quarantine in the pediatric population consisted of a growing and widespread state of anxiety, fear, anger, and uncertainty [2]. Adolescents affected by already diagnosed psychological and neuropsychiatric disorders were at higher risk of flaring up their condition or developing new mental health disorders [3, 4].

According to the Diagnostic and Statistical Manual of Mental Disorders-5 (DSM-V), Somatic symptom disorder (SSD) is a diagnosis that describes a cluster of patients who have distressing somatic symptoms, abnormal thoughts, feelings, and behaviors leading to disruption and distress of daily functioning [5]. SSD has a 10% prevalence in adolescence, coexisting with anxiety and depression in about 30% of patients [6], affecting patients and families’ quality of life causing school absenteeism [7], separation from their peers, and repeated medical evaluations or hospital admissions [8].

An obligated quarantine can be considered as a forced and unusual experimental setting. Studying its impact and understanding the reactions of patients and parents can help identify protective or adverse factors that play a role in SSD. This may help in further recognizing the role of different triggers and dynamics in developing treating strategies or better defining possible risk factors.

In a previous study, we investigated the impact of the national lockdown on psychological symptom burden in a cohort of Italian adolescents with SSD, demonstrating an improvement in depression and anxiety during the forced isolation period by using the Children’s Depression Inventory Short Form (CDI-2-SF) and the Multidimensional Anxiety Scale for Children Self Report (MASC-2-SR) questionnaires, respectively [9]. In particular, the T-score physical symptom domain of the MASC-2-SR was significantly lower in SSD adolescents when compared to non-SSD peers.

This further study aimed to examine the attitudes, behaviors and coping strategies of the same cohort of adolescents with low burden of disease, including the relationship with peers and parents.

This research was part of a cross-sectional observational study conducted in May 2020 at the tertiary level, university teaching, children’s hospital, Institute for Maternal and Child Health of Trieste, Italy. Eligible participants were adolescents aged between 13 and 18 years, experiencing the Covid-19 pandemic restrictive measures.

The inclusion criteria were: (1) adolescents who received a diagnosis of SSD at the Institute, within the previous one year, confirmed by a neuro-psychiatric specialist according to the DSM-V, (2) youths who previously accessed the Institute for an acute injury or an acute organic disease in the same period, matched for age and sex. The exclusion criteria were considered the inability to understand the Italian language, the presence of cognitive impairment, a chronic disease, a neuropsychiatric disorder, or any history of non-investigated chronic pain.

All the adolescents of the SSD groups were clinically evaluated jointly by a trained paediatrician and a pediatric psychiatrist.

Adolescents in the control group were consecutively enrolled if admitted for acute trauma (fractures) or illnesses (appendicitis, testicle torsion, pneumonia) without any referred previous history of chronic illness or psychological problem.

The SSD diagnosis was based on a clinical interview conducted by a child neuropsychiatrist and symptom ratings according to the DSM-V criteria [5]. The electronic medical records of all patients admitted between May 2019 and May 2020 were reviewed to collect information about age, gender, symptoms, symptoms complained at admission, diagnostic test, specialist consult, previous hospitalizations, and final diagnosis, including SSD, at discharge.

Based on the defined criteria, we identified two groups:

A letter explaining the purpose of the study was sent to parents of adolescents who met the inclusion/exclusion criteria. Written informed consent was signed before participation. An anonymous semi-structured SurveyMonkey questionnaire was offered (see online Supplementary file 1) to both adolescents’ groups. It included a general demographic survey, a modified version of the Health Questionnaire Physical Symptoms 15 (PHQ15), the MASC-2-SR and the CDI-2-SF questionnaires. Data regarding the MASC-2-SR and the CDI-2-SF have already been published in another study analyzing depression and anxiety tendencies [9].

The general demographic survey included 18 questions with binary, multiple choices, and the possibility of providing open answers. For each enrolled patient, the following demographic variables were collected: age, gender, information on coping strategies during adolescents’ free-time, including the following main domains: ‘purpose of internet usage,’ ‘pattern of social media, videogames and other leisure usages,’ ‘relationships with peers and parents.’ We also added two questions about general worries and any perceived positive aspect about the ongoing pandemic period to describe adolescents’ general coping strategies during the quarantine period. The PHQ15 questionnaire was administered with the intended goal of quantifying the somatic symptom burden perceived by this cohort of patients [10]. This questionnaire, which investigated somatic symptoms or symptom clusters that accounted for over 90% of the physical complaints reported in the outpatient setting [11], was initially composed of 15 items scoring 0–2 for each. In this study it was administrated with 14 items, excluding the one strictly related to the sexual sphere. According to previous studies with the PHQ-15 [12], missing values were replaced with the mean value of the remaining items if the number of absent items was less than 20%. If the number of missing items in the scale exceeded 20%, the sum score was not computed and counted as missing. The overall results over 14 items were then weighted to 15 items, to compare the result of this study to other previous studies and reference ranges (see Supplementary file 2). The total PHQ-15 score ranges from 0 to 30, and scores of ≥ 5, ≥ 10, ≥ 15 represent mild, moderate, and severe levels of somatization [10]. To assess the anxiety tendency we used the MASC 2-SR which quantifies adolescents’ anxiety across ten domains: separation panic (SP), generalized anxiety disease (GAD), humiliation and refusal (HR), performance fear (PF), social anxiety (SA:T), obsessive compulsive (OC), panic (P), tension and restlessness (TR), physical symptoms (PS:T) and harms avoidance (HA). It investigates 50 items with a score from 0 to 3 for each item. T-scores are categorized into six classifications: Very Elevated (score 70 +), Elevated (score 65–69), Quite Elevated (score 60–64), High Average (score 55–59), Average (score 40–54), and Low (score < 40) [13].

The primary study outcome was to detect differences in attitudes and behaviors during the COVID-19 lockdown, including coping strategies and relationships with parents and peers, between SSD group and their healthy peers. Ethics Committee and approval number: IRB, RC 10/20.

The sample size of 94 (47 for each group) was predetermined to carry out the study, assuming a between groups minimum clinically significant difference of 5 in the mean MASC 2-SR total score.

G* power software, two sided Wilcoxon-Mann Whitney test, 2 groups).

Categorical data were presented as number and percentage, continuous data as median and interquartile range (IQR). Differences between groups (i.e., male vs female or healthy vs SSD) were evaluated with the chi-squared test for categorical data and with the non-parametric Mann–Whitney test for continuous data. The use of nonparametric tests is justified by the non-normal distribution of data, evaluated both visually and with the Kolmogorov–Smirnov test. Differences with p-value < 0.05 were considered statistically significant. Analysis was performed using SPSS version 23 (IBM, New York, USA).

The survey was offered to 160 teenagers. Forty-five subjects (28.1%) refused (22 in the SSD group and 23 in the control group) and 115 were enrolled, 58 (50.4%) with SSD (mean age 15.3 years, 48.3% males), and 57 (49.6%) control peers (mean age 15.8 years, 54.4% males). All subjects had been home-confined for nine weeks when they completed the survey.

This observational study shows that, after nine weeks of COVID-19-related home isolation, when compared to a control group of peers, a broader proportion of adolescents with SSD were more likely to have their relationship with peers and parents improved during the quarantine. This experience is of particular interest when considering the vital link between peers and parents’ relationship and SSD development in the long term period [14]. In the context of pandemic restrictions, the lockdown may have acted as a social relief for most adolescents with SSD by reducing competition with peers and family expectancies, as demonstrated by the low burden of physical symptoms, the hallmark of this condition, as found by the PHQ-15 and MASC-2SR questionnaire. On the other side, social avoidance is strongly correlated with SSD [15], and the comfort reported by patients may have been related to fewer physical symptoms [16].

One may speculate that home confinement could be perceived as a substantial stressor for healthy people and, conversely, as a relief for SSD patients. Although SSD subjects could have experienced reduced external stressors, they may also, on the contrary, have benefited from improved relationships with peers and adults. In this context of school closure and limitation of social relationships, adults' expectations were reduced, performance pressure was limited, and peer competition could be less evident, while solidarity could be increased. This study shows a trend toward a better relationship between SSD adolescents and their parents during isolation compared to healthy peers, thus suggesting a possible protective parental role. However, parents’ s relationship could also take on regressive connotations such as the constant presence of the parent, the continuous attention on the adolescent, reduced only by interference such as school and work.

In this extreme pandemic contest, the condition that characterized the phases of somatic symptoms of a psychogenic nature of adolescents was therefore recreated: dedicated attention and avoidance of sources of stress.

On this topic, albeit in absence of statistical significance, a different Internet pattern of usage was detected in the two groups, being less switched and with less total average time spent to the social media utilization and more switched to the videogames utilization in the SSD group when compared to the non-SSD group. This pattern may in part explain the better self-esteem trend in SSD-adolescents [9], as less involved in comparison with peers, as detailed by the upward social comparison and reinforcing spirals hypotheses and the displacement hypothesis [17].

On the other hand, the presence of intense virtual activity, which facilitated relationships, could reproduce some dynamics of conflict and competition among peers, explaining the similar proportion of adolescent with stable relationship with peers between the two groups [18]. Finally, a different role of social media use could be postulated due to its relevant role in adolescents with chronic illness [19]_. It can be assumed that children with SSD were more comfortable during isolation in the context of remote school lessons and more intense use of social media than peers with a different lifestyle, more focused on outdoor activities and face-to-face relationships, thus explaining the higher proportion of patients having improved their relationships with peers.

A further possible explanation for the well-being of the SSD group may also be related to limited access to medical care, which could increase anxiety and confusion in the absence of a correct diagnosis or through repeated visits with no valuable results [20]. The pandemic may have also shifted the focus from somatic symptoms to the viral infection threat, which is well known for having a minor impact on young people.

This study has some limitations. First, PHQ-15 was not determined before the pandemic began, so it was difficult to determine its effect on this condition burden and it was not administered to the non-SSD group since it was not conceived for healthy population. However, the first part of the study in the same cohort, aimed at determining depressive and anxiety tendencies comparing the two groups, showed a reduction in physical symptoms in adolescents with SSD compared to the control group [9]. In particular, according to the PS:T domain of the MASC 2-SR T-score, SSD patients also reported significantly less physical symptoms burden when compared to the adolescents belonging to the control group (p < 0.05). The second limitation was that the study involved a limited number of patients and that the questionnaire has not yet been administered after the lockdown, even if we continue to follow up both groups to assess anxiety and depression tendencies. Finally, a questionnaire performed at a particular time in an adolescent’s history can be subjected to a selection and recall bias.

The study also had some points of strength. One was the accurate definition of the SSD diagnosis, always made jointly by a pediatric psychiatrist and a pediatrician in a group of children in regular follow-up. A further element was the quality of the PHQ-15 and MASC2-SR validated questionnaire, exploring different somatic domains.

The interpretation of these results must be cautious; the isolation lasted about nine weeks in Italy, and more extended periods can lead to different results. Furthermore, physicians will have to pay close attention to the reopening of schools, which could put a rebound pressure on these patients. In this perspective, we suggest that specifically tailored strategies should be prepared and developed to help the adolescent cope with the return to normal activities.

Adolescents with SSD and low burden of physical symptoms during COVID-19 related isolation reported a better relationship with peers and parents when compared to the control group. A deeper understanding of the relevance of specific stress triggers offered by the unusual and highly specific pandemic contest can facilitate understanding the treatment of these patients.