Health-related quality of life, glycaemic control, lifestyle characteristics and SARS-CoV-2 prevalence in children with type 1 diabetes during the COVID-19 pandemic: results of a longitudinal, prospective single-centre Swiss study

Abstract

Background

This study examines the longitudinal effects of COVID-19 pandemic regulations on health-related quality of life (HrQoL), glycaemic control and lifestyle characteristics and describes the epidemiology of COVID-19 in children with type 1 diabetes (T1D).

Methods

This monocentric, prospective study included, from May to November 2020, pediatric outpatients up to 18 years with T1D. During the first 16 weeks, COVID-19–associated symptoms were assessed weekly, while HrQoL, COVID-19 affectedness and appraisal, screen time, and physical activity were assessed monthly. Body-mass-index standard deviation scores (BMI-SDS), HbA1c, and glycaemic monitoring data were collected at clinical visits three monthly. All parameters were reassessed at final visit after 10 months and assigned to the pandemic restriction phases. Estimates of correlation (est) were tested by a linear mixed-effects model.

Results

55 children with T1D were included, with a median age of 11 years and 56.4% male. HrQoL remained normal and stable throughout the pandemic, but showed significant associations with physical activity (est 3.63, p < 0.01), screen time (est − 1.66, p < 0.01) and financial situation (est − 9.15, p < 0.05). During the pandemic, BMI-SDS increased (p < 0.01), but mean HbA1c remained at about 7.5% and glucose time in range improved (p < 0.01). During the lockdown phases, screen time tended to be highest (p < 0.1), while physical activity was lowest (p < 0.01). At baseline, one of 55 patients (1.8%) was COVID-19 seropositive from infection. At the study’s end 20.5% were positive from infection, and 4.9% from vaccination.

Conclusions

In children with T1D, HrQoL and metabolic control remained stable under pandemic regulations, but BMI-SDS and physical activity worsened. Continuation of outpatient care and opening of schools may have contributed to a healthy lifestyle and resilience.

Trial registration

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12887-025-06375-0.

Background

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes coronavirus disease 2019 (COVID-19), which led to a pandemic between 2020 and 2022 [1]. The true incidence rate of this disease is still uncertain, as a high proportion of asymptomatic infections have been reported [2, 3]. Children seem to have better prognoses regarding COVID-19 than adults [3, 4]. Nevertheless, the pandemic, with its accompanying restrictions, has been linked to behavioral and emotional problems in healthy children [5–8], as well as children with chronic diseases, such as type 1 diabetes (T1D) [9]. In an Indian group of adolescents and young adults with T1D, HbA1c levels were correlated with increased psycho-social stress during the pandemic, as well as reduced blood-glucose-monitoring frequency and inadequate physical activity [9]. In contrast, Polish data have demonstrated a stable metabolic control despite weight gain during the pandemic [10].

At the same time, it was known that impaired mental health of parents, mothers in particular, not only affects the mental health and quality of life of children in Switzerland during the COVID 19 pandemic [11], but also predicts the glucose management of children with diabetes in general [12].

The main aim of this study was to investigate whether the COVID-19 pandemic impacted the quality of life and health, including diabetes control, of children with T1D either through infection or pandemic-related restrictions. Specifically, the study examines health-related quality of life, glycaemic control, weight status, physical activity, screen time, nutritional habits, and SARS-CoV-2 infection in a cohort of children with T1D, and longitudinally explores the associations between these pandemic-associated factors. To allow for comparisons with national data, various covariates were selected based on other current Swiss pediatric studies.

Methods

Study design

The ECOPED Study is a monocentric, prospective multicohort study conducted at the Children’s Hospital of Eastern Switzerland in St. Gallen (OKS) from April 2020 to May 2022. The ECOPED Study consists of three cohorts, namely pediatric outpatients with T1D, a group of healthy children participating in the prospective birth cohort Childhood Allergy, Nutrition, and Environment (CARE) study, and a cross-sectional cohort of pediatric inpatients at the children’s hospital who provided general consent. Only the results for the T1D cohort are reported here.

Children and adolescents attending the outpatient clinic with T1D, aged up to 18 years at study enrollment, were included. Ethics committee approval and written informed consent were obtained as indicated below. Either the children and/or their parents had to have the ability to complete German language questionnaires. The exclusion criteria were the inability to sign the informed consent or the withdrawal of consent. Fifty-five of 173 patients up to 18 years from the diabetes outpatient clinic agreed to participate. The first clinical visits took place between May 2020 and November 2020. The study design is shown in detail in Table S1 (see supplementary file). All visits were combined with 3 monthly scheduled routine controls in the outpatient clinic of diabetology and conducted by a diabetologist. The standard deviation score for body mass index (BMI-SDS), which was calculated from references used in Switzerland [13], was determined at all visits. In the first 4 months, participants completed weekly questionnaires on COVID-19-related symptoms, as well as monthly online questionnaires on quality of life and lifestyle characteristics, namely physical activity, screen time, and COVID exposure and appraisal). At the first and final visits, retrospective assessments were conducted regarding the nutritional composition of the children’s diets, as well as parental mental health and significant life events. Data on Glycaemic control (HbA1c, average glucose (AVG), and time in range (TIR)) were extracted from the laboratory information systems and diabetes-management applications described below.

To measure developments over time, the data collected were assigned to the various pandemic phases in Switzerland [14], as indicated in Table 1. Baseline data from before the pandemic (Phase 0) were collected retrospectively.

Table 1.

Phases of the COVID-19 pandemic in Switzerland

Phase	Time	label	content
0	4 months to 15. March 2020	Pre-COVID	No influence due to the pandemic. For retrospective metabolic data, the time frame was limited to four months prior to the first lockdown.
1	16. March 2020–10. May 2020	First Lockdown	Closed schools, shops, restaurants, and social activities
2	11. May 2020–18. October 2020	Normal	No restrictions
3	19. October 2020–21. December 2020	Slowdown	Assembly restrictions and early closing hours
4	22. December 2020–28. February 2021	Second Lockdown	Closed shops, restaurants, and activities but open schools
5	01. March 2021–31. May 2021	normal	No restrictions. Endpoint is defined as the end of the study

Laboratory parameters

Serum anti-S and serum SARS-CoV-2 nucleocapsid protein immunoglobulin (anti-N) antibody levels were determined using the Elecsys^® electrochemiluminescence immunoassay (ECLIA) (Roche Diagnostics, Rotkreuz, Switzerland) conducted with a COBAS 6000 instrument [15, 16]. Seropositivity was defined using a cutoff of ≥ 0.80 binding antibody units (BAUs)/ml for serum anti-S and a cutoff index (COI) of ≥ 1.0 for serum anti-N, according to the manufacturer’s recommendations.

HbA1c was measured from capillary blood by Alere Afinion™ 2 [17]. Glucose profile calculations with Time in Range (TIR, percentage of time the glucose is within the range 3.9–10.0 mmol/l) and average glucose (AVG) were done using one of the following evaluation programs: Diabass^® version 6 (mediaspects GmbH, D-72336 Balingen), CarelinkTM (Medtronic Diabetes Inc., Northridge, CA 91325 − 1219, USA), Libre View Version 3.10 (Abbott Diabetes Care Ltd., Witney, Oxon, OX29 OYL, UK), or Dexcom Clarity Clinic Portal (Dexcom International Switzerland, CH-6048 Horw). For AVG, data from continuous interstitial glucose monitoring (CGM) and self-monitored capillary blood glucose (SMBG) were used. For TIR calculations, only CGM data were included.

Glycaemic control data were assigned to pandemic phases as follows: for phase 0 (Table 1), available data from 01. December 2019 to 15. March 2020 were used. For the other phases, data were extracted from monitoring systems covering the full duration of each phase, except for phase 2, where data extraction was limited to the 3 months with the highest available device data. The HbA1c values measured in the first 2 weeks of a phase were assigned to the prior phase because 50% of HbA1c was formed in the 4 weeks prior to sampling, so the most recent glycaemic levels have the highest impact on HbA1c [18].

Questionnaires

All questionnaires were completed via the web using the Jotform (www.jotform.com) application, with data being sent via a secure-sockets-layer-encrypted form and servers located in the EU. Answers to the food frequency questionnaire were collected on paper. In the case of children younger than 8 years, the questionnaires were completed by a parent. The questionnaires on parental mental health and life events were completed by a parent.

Health-related quality of life concerning physical health, emotions, leisure, family, friends, and school was assessed using the Kidscreen-10 questionnaire, validated for german language [19, 20]. The results were calculated with an SPSS syntax provided by Ravens-Sieberer [19] and adapted and validated for the Swiss pediatric population [20]. A T-value of 50 represents the mean HrQoL of the German-speaking Swiss pediatric population, and higher or lower T-values indicate a better or worse HrQoL, respectively. Simultaneously, affection throughout and the appraisal of the pandemic were assessed using items kindly provided by Schmidt et al. [11] (Table S2).

The questions about screen time and physical activity [21] were kindly provided by Prof. Susi Kriemler (Epidemiology, Biostatistics & Prevention Institute, University of Zurich). The answers were used to calculate the mean time spent on physical activity during the week and on weekends, as well as the total screen time (phone, TV, and computer games) per day. The questions are based on a US-American study [22].

Parents’ mental health was assessed using a validated nine-question screening tool for the presence and severity of depression known as the Patient Health Questionnaire-9 (PHQ-9) [23, 24]; the validated Scale for General Anxiety Disorder-7 (GAD-7), encompassing 7 questions [25, 26]; and the Life Events Scale (Table S3) [27], in which the number of reported life events was summed for analysis.

The macro- and micronutrient intake of the children (carbohydrates, proteins, fats, and fiber) was calculated based on the intake of various food groups, with amounts and frequencies being reported on the validated Food Frequency Questionnaire (FFQ) [28, 29]. Intake was evaluated with the German-Austrian-Swiss (D-A-CH) reference values [30] for each age group. Questionnaires were excluded if there were unrealistic values, specifically if the total daily calorie intake was lower than 70% of the lowest recommended calorie intake.

The questionnaires about COVID-19-related symptoms (Table S4) and the completion questionnaire (Table S5) on external SARS-CoV-2 tests, isolation, quarantine, SARS-COV-2 vaccination status, and anxiety about COVID-19 were validated.

Statistical analysis

Descriptive statistics were reported for all phases in the form of sample size (n), arithmetic mean, and standard deviation for continuous variables and percentages for categorial variables. For test statistics, the level of significance is defined as p < 0.05, and a p value < 0.1 is defined as a trend. The primary endpoint is HrQoL. Secondary endpoints, such as BMI-SDS, and glycaemic control (HbA1c, AVG, and TIR), physical activity, screen time, were analysed accordingly. In cases in which a parameter was measured more than once for a given phase of the pandemic, the phase-specific averages were used for further calculations.

Test statistics intended to analyse significant changes between the phases, considering the individual slopes of each participant, were used with a linear mixed-effects model (with a type III Analysis of Variance Table and Satterthwaite’s method), with phase as a fixed effect and patient-ID as a random effect. Because the method does not exclude cases with missing data, all available data will be included in this analysis.

To assess the correlation between HrQoL and various continuous or dichotomous parameters, a linear mixed-effects model was run, with HrQoL as an outcome, patient-ID as a random effect, and the following predictors: physical activity, screen time, the child’s concern about being affected by COVID-19, the child’s concern about their family being affected by COVID-19, financial difficulties due to the pandemic, and family members being part of at-risk populations. Likewise, to assess the correlation between glycaemic control and various parameters, a linear mixed-effects model was run with glycaemic control as the outcome (HbA1c, AVG, and TIR), patient-ID as a random effect, and the following predictors: health-related quality of life, physical activity and screen time. The same was done for weight status (BMI-SDS). All linear mixed-effects model correlations are reported with estimates. The association between HrQoL and physical activity can be used as an example of the association between continuous predictor variables: a significant association with an estimate of 3.63, indicates that HrQoL increases by 3.63 units for every increasing unit of the predictor variable (physical activity in hours per day). The association between HrQoL and “financial affectedness throughout the pandemic” can be used as an example of dichotomous variables: A significant association with an estimate of −9.15 indicates that HrQoL decreases by 9.15 units for every positive expression of financial affectedness throughout the pandemic. The associations between the primary and secondary endpoints and parameters that were only assessed at study’s beginning and end (age, GAD-7, PHQ-9, and life events) were tested using Spearman’s rank correlation in each phase. This was also applied to the differences between the phases (delta values). Spearman’s rank correlation results were adjusted for multiple comparisons using the Benjamini-Hochberg method to reduce false positives. Spearman’s rho and p-values are reported for significant results. The level of significance is set at 5% (p < 0.05).

Results

Baseline characteristics

A total of 55 children and adolescents with diabetes were included. At baseline (Table 2), 56.4% of patients were male, and 27% had a migration background, (i.e. at least 1 parent born outside Switzerland or the European Union). The median values in terms of age, BMI-SDS, and HbA1c were 11 years, 0.55, and 7.45%, respectively. Regarding glucose monitoring types, self-monitored capillary blood glucose (n = 10) and CGM (n = 45, including flash glucose monitoring) were used. As an insulin application technique, multiple daily injection (MDI) therapy via pen (n = 18) and continuous subcutaneous insulin infusion pump therapy (n = 37) were used. Six participants wished to discontinue within the first 4 months and did not complete any additional questionnaires, but based on their additional general consent provided beforehand, their metabolic data from the diabetes outpatient visits could still be included in the study.

Table 2.

Characteristics of study population at baseline¹

	n or %	mean	median	min	max
Patients with diabetes	n = 55
Sex	56.4% male
Age (years)	n = 55	11.1	11.0	4	18
BMI-SDS	n = 51 2	0.51	0.55	−1.52	2.79
HbA1c (%)	n = 50 3	7.70	7.45	6.50	11.9
AVG (mmol/L)	n = 48 4	9.34	7.20	5.60	13.7

BMI-SDS body mass index–standard deviation score, HbA1c glycosylated hemoglobin, AVG average glucose

¹Data from Phase 0 are used as a baseline

²Missing data are due to patients not yet being diagnosed with diabetes at phase 0 (n = 4)

³Missing data are due to patients not yet being diagnosed with diabetes at phase 0 (n = 4) and one child with no available HbA1c values

⁴Missing data are due to patients not yet being diagnosed with diabetes at phase 0 (n = 4) or the fact that no glucose sensor data were available (n = 3)

Primary endpoint parameter: HrQoL

Health-related quality of life scores fluctuated around the mean of the normal ranges, with no significant change throughout the phases (p = 0.27). Phase 4 was excluded from the analysis due to limited data (n = 7). There was a significant negative correlation between HrQoL and age in phase 5 (r = −0.51, p = 0.04) indicating that HrQoL was lower in older children at studies end; a significant positive correlation with physical activity (est = 3.63, p < 0.01) and a significant negative correlation with screen time (est = −1.66, p < 0.01 and financial affectedness throughout the pandemic (est = −9.15, p = 0.043), HrQoL was negatively correlated with parents’ initial GAD-7 scores (delta HrQoL [phase 5 - phase 1], r = −1.0, p < 0.01), meaning that HrQoL increased more during the pandemic when parents’ anxiety scores were lower at beginning.

Secondary outcome parameter

Tables 2, 3 and 4 and S6 to S8 in the Supplementary Material show the results of secondary endpoints and their associations.

Table 3.

Results of primary and secondary outcome parameters

	0 pre-COVID	1 lock-down	2 normal	3 slow-down	4 lock-down	5 normal	p-value1
HrQoL (T-value)							0.27
n		11	49	13	7	35
Mean		46.69	50.58	48.92	52.633	51.13
SD		7.21	8.31	8.92	11.15	9.24
% of basal mean		100	108.33	104.78	112.72	109.51
BMI-SDS							< 0.01
n	51	44	54	44	43	49
Mean	0.51	0.62	0.66	0.69	0.87	0.82
SD	0.92	0.99	0.95	0.98	1.02	0.99
% of basal mean	100	121.57	129.41	135.29	170.59	160.78
HbA1c (%)							0.07
n	49	43	53	40	43	53
Mean	7.70	7.46	7.64	7.72	7.69	7.49
SD	1.00	0.85	1.17	1.26	1.02	1.06
% of basal mean	100	96.88	99.22	100.26	99.87	97.27
AVG (mmol/l)							0.63
n	48	48	54	51	47	46
Mean	9.34	9.26	9.39	9.45	9.19	9.14
SD	1.82	1.67	1.85	1.62	1.66	1.68
% of basal mean	100	99.14	100.54	101.18	98.39	97.86
TIR (%)							< 0.01
n	41	41	45	44	42	40
Mean	52.7	55.02	53.94	55.12	59.15	60.02
SD	13.86	12.71	12.88	15.22	12.78	14.18
% of basal mean	100	104.40	102.35	104.59	112.24	113.89
Screen Time (min/d)							0.09
n		43	49	13	7	35
Mean		187.21 9	148.13	169.23	208.933	175.29
SD		9	111.77	119.51	108.20	125.70
% of basal mean		100	79.12	90.40	111.60	93.63
Physical Activity (h/d)							< 0.01
n		38	43	13	7	35
Mean		1.09	1.78	1.34	1.213	1.44
SD		0.68	0.91	1.05	0.76	1.19
% of basal mean		100	163.30	96.33	111.01	132.11

HrQoL Health-related quality of life, BMI-SDS body mass index–standard deviation score, HbA1c glycosylated hemoglobin, AVG average glucose, TIR time in range, n number of children with available data per phase, SD standard deviation

¹Linear mixed effects model with Type III Analysis of Variance Table with Satterthwaite’s method, boldface indicates significance, if < 0.05, or a trend, if between 0.05 and 0.1

³These data were not included in the test of significance, because of the low available data

Table 4.

Associations between health outcome measures and COVID-19-related influencing factors

	HrQoL1	Physical activity 1	Screen time1	Childs concern to be affected by COVID-191	Childs concern for family to be affected by COVID-191	Financial affectedness through the pandemic1	Family members in risk population1	Age2	GAD-72		PHQ-92		Life Events2
									begin	end	begin	end	begin	end
HrQoL		3.63**	−1.66**	n.s.	6.41**	−9.15*	4.06*	Phase 5 −0.51*	Δ5 − 1 −1.0**	Phase 2 −0.51*	n.s.	n.s.	n.s.	n.s.
BMI-SDS	n.s.	n.s.	0.07**					n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	Δ1-0 −0.53*
HbA1c	0.02*	−0.17*	n.s.					n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.
AVG	n.s.	n.s.	n.s.
TIR	0.27 t	n.s.	n.s.
Physical activity								Phase 2 −0.45 t	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.
Screen time								Phase 1 0.70 ** Phase 2 0.62** Phase 3 0.76* Phase 5 0.55**	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.

HrQoL Health-related quality of life, BMI-SDS body mass index–standard deviation score, HbA1c glycosylated haemoglobin, AVG average glucose, TIR time in range, PHQ-9 nine-question patient health questionnaire used as screening tool for depression, GAD-7 seven-question generalized anxiety disorder as screening tool for a generalized anxiety disorder, n.s. not significant

^t = p < 0.1 (trend)

* = p < 0.05, ** = p < 0.01

¹Estimate of correlation using a linear mixed effects model comparing individual slopes during all phases. Reported if p < 0.1. Refer to results for further explanation

²Coefficient r of Spearman’s rank correlation with Benjamini-Hochberg procedure is reported if p < 0.1 for each timepoint, as well as the differences between the phases, which are labeled as delta (Δ)

BMI-SDS and glycaemic control

Throughout the pandemic, BMI-SDS significantly increased (p < 0.01, Table 3), with the highest mean appearing in phase 4 (0.87), and it was predicted by screen time (est = 0.07, p = 0.004, Table 4).

In the assessment of glycaemic control, four data points with outlier values were excluded due to newly diagnosed diabetes. Glycosylated hemoglobin values showed a trend with most favorable values during the lockdown (mean = 7.46%, SD 0.85). Glycosylated hemoglobin levels showed a significant positive correlation with HrQoL (est = 0.02, p = 0.014) and a negative correlation with physical activity (est = −0.17, p = 0.026).

Time in range based on CGM increased during the pandemic (p < 0.01), with the lowest mean being observed before the pandemic (52.7%) and the highest mean being observed in phase 5 (60.02%; Table 3). There was no correlation with other parameters (Table 4).

Lifestyle behaviors

Daily physical activity significantly increased after the lockdowns and over the entire study. It was highest after the first lockdown (p < 0.01, Phase 2, 1.78 h, Table 3) and negatively correlated with age (Table 4; i.e., it was higher in younger children).

Mean daily screen time was highest during lockdowns (187 and 209 min) and showed an increasing trend until the end of the pandemic (Table 3). Throughout most of the pandemic phases, screen time was positively correlated with age (see Table 4).

Nutrient intakes according to the FFQ (Table S6), which was mostly completed by the mothers, were mainly within the German-Austrian-Swiss reference values [30]. Notably, the daily energy intake from fats and proteins was generally above the recommendations. Due to the low number of questionnaires per age-related reference group, a detailed evaluation of nutrition before and after the pandemic was not possible.

Parental health and appraisal of COVID-19

The number of parents with mental health scores indicating severe issues and the mean PHQ-9 and GAD-7 scores were higher after the pandemic than at its beginning (Table S7). The concerns of a child about being affected by COVID-19 and their concerns for their family members decreased throughout the pandemic (Table S7). The concern of the child for their family members was generally higher than the child’s concern about themself being affected (Table S7). The separate evaluation of questionnaires completed by children over 8 years of age or via proxy assessment for younger children revealed no differences, so family concerns were not influenced by over-assessment bias created by the parents responding instead of the child.

SARS-CoV-2 antibody tests, infections and survey on symptoms and vaccination

At study’s beginning, one of 55 tested participants were SARS-CoV-2 antibody positive due to infection (1.8%, SARS-CoV-2 antibody positivity without a history of vaccination). At the study’s end, six of the 41 tested participants were seropositive due to infection (14.6%; five persons were SARS-CoV-2 IgG positive without a history of vaccination at sampling and one person was SARS-CoV-2-S1-RBD-antibody and SARS-CoV-2-N-antibody positive). In the final questionnaire, 16 participants reported having had a COVID-19 test outside of this study, of which four were self-reportedly positive and one was also tested positive within the study. Considering both in-study and out-of-study testing, nine of 44 available results (20.5%) were SARS-CoV-2 positive.

At study’s end, two of 41 were seropositive after vaccination (4.9%; one person was SARS-CoV-2-S1-RBD-antibody positive and one person was SARS-CoV-2 IgG positive with a history of vaccination). One participant who was IgM seropositive due to infection later during the study was again seronegative at the study’s end.

At the end of the study, the completion questionnaires were answered by 39 participants only, therefore, the survey on COVID-19 symptoms, contacts, and vaccination status is presented in the Supplementay Material.

Discussion

This study is the first clinical prospective study to longitudinally report on quality of life, glycaemic control, weight status, lifestyle characteristics, and COVID-19 seroprevalence in a Swiss cohort of children with diabetes, in the specific context of the healthcare situation in Switzerland during the pandemic. Even during lockdown, outpatient care for pediatric diabetes patients and the supply of medical goods, such as medications and medical devices, were not restricted.

Children with T1D were not particularly affected by COVID-19, and their HrQoL did not systematically deteriorate during the pandemic. Irrespective of HrQoL levels, children with diabetes did not perceive themselves to be at an increased risk, especially at the end of the study (Table S7). Indeed, the Swiss Society of Pediatric Diabetology had propagated the statement that T1D in children and adolescents does not represent an increased risk for severe forms of COVID-19 disease [31]. However, regular diabetes consultations should continue in outpatient clinics throughout the pandemic. Although physical activity did significantly worsen during lockdowns and BMI-SDS did significantly increase during the pandemic, glycaemic control and other lifestyle behaviors did not worsen, as will be discussed in more detail below.

In this study, findings indicate that children with T1D in the German-speaking region of Switzerland maintained stable HrQoL throughout the pandemic phases. This stands in contrast to the significantly reduced HrQoL observed in the pediatric population in Germany during the pandemic, with that nation’s considerably longer school closures and lockdown phases [5]. Compared to a pre-pandemic Swiss study [32], the mean T-values observed in this study were consistently lower across all phases than those reported for healthy Swiss children (T = 52.8 ± 9.23) and were also lower during lockdown and slowdown phases than those reported for children with obesity prior to the pandemic (T = 49.01 ± 8.66).

A higher HrQoL was associated with higher physical activity and lower screen time, confirming the results of the Swiss “Ciao Corona” Study, which showed higher life satisfaction [21] and HrQoL [33] in children fulfilling the recommended physical activity time and screen time during the pandemic. A higher HrQoL was also associated with lower financial affectedness through the pandemic which is, again, in line with the results of another study. Schmidt et al. [11] found, for children in German-speaking Europe, more behavioral and emotional problems in those affected by COVID-19 or its negative appraisal. A decrease in HrQoL during the pandemic was associated with higher parental anxiety scores prior to the pandemic, suggesting that children may have had reduced resilience or coping capacity in response to pandemic-related restrictions when their parents experienced elevated anxiety beforehand.

In terms of glycaemic control, HbA1c remained stable, and TIR significantly improved despite several impairments due to pandemic restrictions, as further discussed below. At baseline, the study group had mean HbA1c of 7.7%, slightly below the published averages for Switzerland and Germany of 7.9% [34].

Similarly, in German data on children with T1D [35], as well is Polish data [10], stable metabolic control during the pandemic, despite increasing BMI-SDS, was shown. One explanation for stable diabetes control may be that despite the pandemic, the children had regular outpatient medical care and faced no supply shortages, as well as having high-quality patient and family training on self-efficacy in the face of chronic disease.

An overall increase in BMI-SDS of 0.31 points was observed over the 16-month period from phase 0 to phase 5, suggesting a potential association with the pandemic-related restrictions. This pattern is comparable to findings from a German study reporting a BMI-SDS increase of 0.24 points during the lockdown year [36]. The increase in the St. Gallen study is higher than the increase of BMI-SDS in children with diabetes prior to the COVID-19 pandemic, described in other studies, ranging between 0 SD [37] and 0.2 SD [38].

In the present study, the general increase in BMI-SDS was associated with increased screen time. A global review [39] also highlighted the increase in BMI and its relationship with reduced physical activity during the pandemic, while the relationship between screen or sedentary time and overweight is already well known [40].

Like the results of the “Ciao Corona” study [21], this study showed a significant low point for physical activity being observed during the first lockdown, and a non-significant change in screen time, with the highest level of screen time being observed during both lockdowns. The course of the pandemic may have exacerbated unhealthy lifestyle habits among adolescents, because, in the present study, older age was associated with longer screen time in almost all phases of the study and with lower physical activity after the first lockdown. Similar associations were also observed in German-speaking Europe, where poorer parental mental health was associated with more emotional and behavioral problems during the pandemic [11].

In the nutritional assessment, the energy intake through fats was, in all times, and nearly all age groups and higher than the recommended ratio by D-A-CH [30]. This could be due to avoiding carbohydrate intake to decrease the number and/or amount of insulin injections. Contrary to this hypothesis, the higher intake of fats could also be a national dietary habit, as shown in Swiss adolescents, whose daily energy intake was composed of 37% fats [41].

Notably, the parental mental health in this study group was better than in the cohort of children in German-speaking European countries: the mean depression score (5.12 for parents of children 1–6 years and 4.44 in parents of children 7–19 years) and anxiety score (4.41, 3.96) were higher in Schmidt et al. Interestingly, the higher the number of negative life events assessed at the end of the study, the less weight they gained during the first lockdown. It is thought that parent’s concerns about the initial uncertainty of the effects of the pandemic led them to reinforce healthy habits.

Only 15% of the parental mental health data in our study were provided by fathers (n = 9). However, given that previous research [11] has indicated maternal—but not paternal—personality traits are more closely associated with children’s glycaemic control, this low rate of paternal participation is considered acceptable. In addition, the responses from fathers were descriptively similar to those from mothers.

An increase in SARS-CoV-2 seroprevalence was observed, rising from 1.8% at baseline to 20.5% at study completion, including self-reported positive test results. A Swiss cohort study [42] reported seroprevalences of 5.6% in June 2020 and 18.4% in April 2021, values comparable to those observed in the present study. No patients were severely affected or hospitalized, and none died from COVID-19. This suggests that the risk of COVID-19 in the cohort of children with diabetes was similar to that in the general pediatric population. Surprisingly, all reported contacts with SARS-CoV-2 in school settings were due to other children, not school staff.

The vaccination rate among participants aged 10–19 years was 29% (eight vaccinated children out of the 27 answers in the corresponding age group), which is lower than the national Swiss rate of 50% for the same age group as of May 2022 [43]. The most common arguments against vaccination were concerns about a too-short research period and side effects. These were consistent with another survey conducted in Switzerland [44], which reported a desire for more evidence and concerns about side effects as the most common arguments against vaccination among parents of children from 5 to 13 years old.

One limitation of this study is the low number of participants as compared to other nationwide studies. However, the characteristics of the study group were roughly similar to those of the entire T1D cohort at our centre, with an average age of 12.8 years and an HbA1c value of 7.5%, with 73% receiving continuous subcutaneous insulin infusion pump therapy and 36% having a migrant background. Nonetheless, the limited sample size reduces the statistical power of the study, potentially obscuring meaningful associations that may no longer reach significance after statistical correction.

The glycaemic data were collected with different sensory methods and included children with different insulin application techniques. Since the aim of this study was not to examine the results of diabetes treatment techniques, but rather the physical and mental health of children with diabetes during the pandemic, this limitation was already taken into account in the planning phase.

Methodologically, the reported SARS-CoV-2-associated symptoms and actual infections were not confirmed e.g. by antigen tests and antibody testing, since these were infrequent. Also by another Swiss study [45], symptom assessment was shown to be insufficient to predict a SARS-CoV-2 infection.

Conclusions

In summary, children with diabetes from German-speaking Switzerland were affected by the pandemic in a similar way to healthy children. The children with T1D included in the present study appeared to benefit from sufficient resilience factors, such as a fairly good quality of life and relatively stable parental mental health. Thus, their HbA1c levels did not worsen despite increasing weight and a less favorable lifestyle during the pandemic. In addition to adequate familial and socioeconomic resources in a selected longitudinal study population, it can only be assumed that the favorable health outcomes were also due to sound medical care throughout the pandemic. As shown in other studies, the children were unduly burdened by the restrictions of the pandemic, leading to lasting deterioration in terms of their health behaviors in the form of increased media consumption and weight gain. In the interest of children’s health, therefore, school closures and restricted access to outpatient care should only be implemented cautiously and when there are no alternatives.

Abbreviations

Anti-N

SARS-CoV-2 Nucleocapsid Protein Immunoglobulin Antibody

AVG

Average Glucose

BAUs

Binding Antibody Units

BMI-SDS

Body Mass Index Standard Deviation Score

CARE

Childhood Allergy, Nutrition, and Environment

CGM

Continuous Interstitial Glucose Monitoring

COI

Cutoff Index

COVID-19

Coronavirus Disease 2019

D-A-CH

German-Austrian-Swiss

ECLIA

Electrochemiluminescence Immunoassay

ECOPED

Epidemiology of COVID-19 in Swiss Pediatric Populations

Est

Estimates of Correlation

FFQ

Food Frequency Questionnaire

GAD-7

Generalized Anxiety Disorder 7-Item Scale

HrQoL

Health-Related Quality of Life

MDI

Multiple Daily Injection

n

Sample Size

n.s.

Not Significant

OKS

Children’s Hospital of Eastern Switzerland in St. Gallen

PHQ-9

Patient Health Questionnaire-9

SARS-CoV-2

Severe Acute Respiratory Syndrome Coronavirus 2

SD

Standard Deviation

SMBG

Self-Monitored Capillary Blood Glucose

T1D

Type 1 Diabetes

TIR

Time in Range

Authors’ contributions

SD, CK, CR, RL, and DL contributed to the study design. SD developed the datasets and online questionnaire tools. Data collection was performed by SD, CK, CR, KH, and DL. SD conducted the data analysis and interpretation, with contributions from CK and DL. SD wrote the initial manuscript draft. All authors critically revised the manuscript and approved the final version. CK, CR, and DL provided project supervision.

Funding

Dagmar l’Allemand: Merck donation # 4501739731/2 for psychosocial care of patients with endocrine diseases.

Data availability

Individual participant data underlying the results presented in this paper will be available following deidentification. The data will be accessible from the date of publication and for a period of five years thereafter. Access will be granted to researchers who submit a scientifically sound and methodologically appropriate proposal. Requests for data access should be directed to serpil.dokhan@gmail.com. Prior to data release, requestors must enter into a formal data access agreement.

Declarations

Ethics approval and consent to participate

The Ethics Committee of Eastern Switzerland (Ethikkommission Ostschweiz) approved the study with the reference number EKOS 20/059. Informed consent was provided by the children and at least one parent before enrolment.

Consent for publication

Not applicable.

Competing interests

Serpil Dokhan-Vural presented early parts of the study at annual meetings of diabetes, endocrine and pediatric societies (Swiss Society of Endocrinology and Diabetology SGED 2021, European Society for Pediatric Endocrinology ESPE 2021, Swiss Society for Pediatrics SGP 2021). Dagmar l’Allemand is co-president of the National Association of Obesity in Children and Adolescents. The other co-authors declare that there are no conflicts of interest regarding the publication of this article.