COVID‐19 related anxiety in children and adolescents with severe obesity: A mixed‐methods study

Summary Recent studies report negative mental health effects of the COVID‐19 related lockdown measures in general paediatric cohorts. Since obesity is a risk factor for COVID‐19 in adults, children (including adolescents) with obesity might perceive themselves to be vulnerable. Using a combined quantitative and qualitative approach, we explored COVID‐19 related anxiety in paediatric patients with severe obesity in the Netherlands using semi‐structured telephone interviews and the Paediatric Quality of Life Inventory (PedsQL) questionnaire, which had also been completed by the study population at baseline in the year prior to the COVID‐19 outbreak. In total, 75 families participated in the semi‐structured telephone interviews during the lockdown, April 2020. Characteristics of included patients were: median age 10.5 years (interquartile range = 7.6‐15.2); 52% female; mean BMI standard deviation score 3.8 (SD = 1.0). COVID‐19 related anxiety was reported for 24/75 (32%) children. The mean decrease in PedsQL score between baseline visit and COVID‐19 outbreak did not differ between children for whom anxiety was reported vs those for whom it was not (mean change −10.3 ± 36.5 vs −3.3 ± 24.4, P = .54). Self‐imposed strict quarantine measures were taken by 19/75 (25%) families. During follow‐up, several families reported that the previous contact alleviated their anxiety. In conclusion, healthcare professionals should address possible COVID‐19 related anxiety in children with severe obesity. Addressing COVID‐19 related anxiety could mitigate its potential negative effects.


| INTRODUCTION
During the current coronavirus disease 2019 (COVID-19) pandemic, governments across the world have used differential lockdown and quarantine measures to mitigate the spread of the virus. Recent studies report how this situation affected the psychological wellbeing of children (including adolescents). [1][2][3][4][5][6][7][8][9] These studies report several adverse effects on psychological wellbeing such as anxiety, worrying, irritability, depressive symptoms and even post-traumatic stress disorder symptoms in 18.9% to 43.7% of children sampled from the Ozair Abawi and Mila S. Welling contributed equally to this study. general population in Asian, European or American countries. Moreover, a recent study in Italian children and adolescents with obesity showed unfavourable changes in eating, sleeping and activity behaviours during COVID-19 quarantaine. 10 Obesity is regarded as a risk factor for COVID-19 in adults. 11 Consequently, children with obesity might perceive themselves to be vulnerable. Moreover, we noticed COVID-19 related concerns during our regular contacts with children and their parents at the outpatient clinic of our paediatric obesity centre when the governmental lockdown measures in the Netherlands were effectuated. On top of that, quality of life is already known to be diminished in children with severe obesity in comparison to the general population. 12,13 However, no studies have assessed such psychological aspects of the COVID-19 outbreak in children and adolescents with obesity. Therefore, we designed a combined quantitative and qualitative study to explore the psychological impact of the COVID-19 outbreak and related lockdown measures in children (including adolescents) with severe obesity and their potential effects on lifestyle behaviour. When conducting this study, COVID-19 related anxiety appeared to be an important theme, similar to results from the previously mentioned literature from general populations. Accordingly, we want to present our in-depth findings regarding COVID-19 related anxiety in children with severe obesity and their parents.

| METHODS
This study was approved by the ethics committee of the Erasmus MC. All data were collected for healthcare purposes and filed in the patient's medical records. Written informed consent was obtained from all patients and/or their caregivers to use their health data for research purposes after pseudonymization.

| Study participants
In the Netherlands, selective lockdown measures including school closures were established from 16 March 2020 onwards. During the first month, between 2 April 2020 and 23 April 2020, when these measures were in full effect, we contacted all parents of children currently under treatment at Obesity Center CGG (Erasmus MC-Sophia Children's Hospital), a national referral centre for obesity.
Patients are referred to Obesity Center CGG for diagnostic evaluation and/or personalized therapeutic advice. 14 We approached parents of all patients who had completed the diagnostic workup of our obesity centre and whose last visit to the outpatient clinic was in 2019 or 2020. We did not approach parents of children who have severe intellectual disability or severe behavioural problems, as we expected that their families' experiences during the lockdown period would not be representative. Because this study was conducted in the context of patient care, we included all eligible study participants even after data saturation for qualitative analyses had been achieved.

| Quantitative assessments and analysis
Height and weight were measured during the previous hospital visit within the past year by trained outpatient clinic assistants and BMI was converted to age-and sex-specific standard deviation scores (SDS) using Dutch reference charts. 15 Both at the baseline visit prior to the COVID-19 pandemic as well as during the lockdown measures, the 23-item Paediatric Quality of Life inventory (PedsQL) 4.0 (parents proxy-report version) was completed. We assessed the total score and the sub-score for emotional functioning, ranging from 0 to 100 with higher scores indicating better quality of life. 16 Quantitative data were analysed using SPSS version 25.0 (IBM). Differences in patient characteristics between patients for whom anxiety was reported compared to those for whom anxiety was not reported in the abovementioned question were analysed using (paired samples) ttests or Mann-Whitney tests with an α of .05.

| Qualitative analysis
Qualitative data were analysed using MAXQDA 2018 (VERBI Software) following best practice methods for qualitative studies and were reported following the Consolidated Criteria for Reporting Qualitative Research (COREQ) checklist. 17,18 Two physicians (OA, MSW) independently coded all interviews according to the Grounded Theory after all telephone interviews had been conducted. 19 According to this theory, first a deductive, theory-driven approach was used, followed by an inductive, data-driven approach, by two of the three interviewing physicians. The two physicians started by open coding of interview data independently. The applied codes were then compared and differences were solved by consensus. Subsequently, a code tree was developed in a meeting with the study team using axial coding.
To minimize the possibility of structural differences between the three physicians who conducted the interviews, the code tree was developed based on interviews from a subset of 24 patients, eight patients per interviewing physician. Finally, selective coding was used to identify the code categories that were most relevant to our research question. The axial and selective coding steps were also performed independently by both physicians and differences were solved by consensus. During the entire qualitative analysis process, a study log was kept by the two physicians and memos were used to carefully note emerging ideas about the data analysis which were discussed during weekly meetings with the study team, to further ensure rigour.

| RESULTS
In total, 90 families were approached. Seventy-five participated in the telephone interviews, of which 40 also completed the PedsQL questionnaire. Table 1 shows the baseline characteristics of the patients.
Anxiety related to the COVID-19 outbreak and related measures was reported for 24/75 (32%) children. Baseline characteristics and quality of life did not differ significantly between patients for whom anxiety was reported vs not reported (Tables 1 and 2). The mean PedsQL total score between baseline visit and COVID-19 outbreak slightly decreased in the study population, although not statistically significant (mean change −6.3 ± 29.9; P = .26). A bigger decrease was seen in the children for whom anxiety was reported vs those who did not (mean change −10.3 ± 36.5 vs −3.3 ± 24.4), but this was also not statistically significant (Table 2). Table 3 10,24 In a significant proportion of the families (25%) in our study, self-imposed quarantine measures were taken, even though measures advised by our national authorities did not differentiate between children with obesity or other chronic diseases and healthy children. These strict self-imposed measures are a concern because they can add to the known negative effects of the COVID-19 pandemic on lifestyle behaviour.
The anxiety that potentially underlies these self-imposed measures seems to be modifiable. In the families for whom short-term follow-up was necessary, we experienced that discussing this emotion with patients and parents and educating them can relieve concerns and make them lift their strict • Child (17 y, F) is afraid that she is more likely to get ill due to Corona because of her health problems. • Child (10 y, M) is afraid he will get more ill than others from Corona. Theme 1.2: anxious for health of family members at risk for COVID-19 due to perceived vulnerability • Child (11 y, M) is concerned for his mother. He always wants to join her during her weekly visits to the supermarket. If it was up to him, she would stay home all the time. • Child (9 y, M) is afraid his father might get ill, because his father has heart failure and COPD.

Theme 2: Reasons for anxiety in parents
Theme 2.1: anxious for child being at risk for COVID-19 due to perceived vulnerability • Mother is afraid that her child (5 y, F) is at increased risk because of her obesity. Therefore, they already confined themselves to home before governmental lockdown measures were taken. • Father is not sure if he will let his son (11 y, M) go to school after school reopenings due to his asthma.
Theme 2.2: anxious for transmitting COVID-19 to family members at risk • Child (15 y, F) is not allowed to have contact with friends, because parents fear she will transmit Corona to their 75-year-old grandfather who lives with them. • Child (11 y, M) is not allowed to play with friends, because of his mother's asthma.
He is also not allowed to visit his grandparents. • Parents cancelled all support and care from healthcare professionals on their own initiative because parents perceive their child (16 y, F) to be vulnerable. • Initially, the family was anxious and stayed at home all the time. Yesterday mother and child (5 y, F) went outside for the first time since 3 wk. • Child (11 y, F) is not allowed to play with friends anymore. Theme 3.2: additional restrictions self-imposed by child only • Child (11 y, M) is afraid to play outside. Even before the national lockdown measures were issued, he declined to go outside when his parents asked him to. In the past 1.5 mo., he only went outside three times. • Child (9 y, M) does not want to meet with friends anymore, because he thinks his father is at increased risk for COVID-19.
Theme 3.2: concerns alleviated by healthcare professional • In the beginning, the child (11 y, F) was afraid to be at risk because of her obesity. After the talk with healthcare professional X her concerns were relieved.

| Strengths and limitations
A strength of our study is our qualitative approach which enabled us to explore possible arguments behind COVID-19 related anxiety and its potential modifiability. Moreover, our relatively large sample size allowed us to reach data saturation. A strength of our quantitative analyses is the comparison of PedsQL scores before and during the COVID-19 outbreak, as it is known that quality of life is already compromised in children with severe obesity. 12,13 A limitation of this study is its cross-sectional analysis; follow-up studies are needed to evaluate the course and effect of COVID-19 related anxiety on weight-related health and will be performed for our patient group.
We did not consider including a control group without obesity because our study was designed to explore the impact of the COVID-19 outbreak and its consequences on lifestyle behaviours specifically in children with severe obesity. Accordingly, our patients served as their own control for the quantitative analyses. This should be kept in mind when attempting to extrapolate our findings.
In conclusion, healthcare professionals should be aware of the possible presence of COVID-19 related anxiety and its behavioural consequences, especially in children with severe obesity. Addressing this anxiety could mitigate its potential negative effects on the psychological wellbeing and lifestyle behaviours of these children.