Retrospective longitudinal study on the long-term impact of COVID-19 infection on polysomnographic evaluation in patients with Prader-Willi syndrome Authors: Sina Braun, Constanze Laemmer, Sandra Schulte, Bettina Gohlke Published: 2024 ABSTRACT AbstractBackgroundTo evaluate the impact of coronavirus disease 2019 (COVID-19) on polysomnographic evaluation in patients with Prader-Willi syndrome (PWS).Patients and methodsA retrospective cohort study of two consecutive overnight polysomnograms (PSG) in 92 PWS patients (mean age 9.1, range 3.1–22 years). 57/92 participants (35 female) had a COVID-19 infection between the two consecutive examinations. 35 patients (21 female) had no infection (control group). Distribution of genetics was as follows: 13/57 (22.8%) deletion, 19/57 (33.3%) uniparental disomy, 2/57 (3,5%) imprinting defect, 3/57 (5.3%) non-deletion, 20/57 (35.1%) diagnosed by analyses of the methylation pattern of chromosome 15q11-13. Mean time interval between COVID-19 infection and post-COVID-19 evaluation was 96.2 days.ResultsCourse of COVID-19 infection was asymptomatic 8/82 (9.8%), mild 63/82 (76.8%), medium 11/84 (13.4%). The five most frequently experienced symptoms in PWS patients were fever (56.1%); headache (45.1%); cold (42.7%); cough (31.7%) and body aches (21.95%). PWS patients who had COVID-19 infection had significantly lower mean oxygen saturation (SpO2) measured by pulse oximetry (post 94.8% vs. pre 95.7%, p = 0.001), lower detected lowermost SpO2 (post 86.2 vs. pre 87.3%, p = 0.003), and higher occurrence of hypopnoea (post 13.9 vs. pre 10.7, p = 0.001). Time in optimal SpO2 (95–100%) decreased significantly (post 54.3% vs. pre 73.8%, p = 0.001), whereas an increase was observed in time in suboptimal SpO2 (90–95%) (post 45.5% vs. 25.8%, p = 0.001) and in time in poor SpO2 (< 90%) (post 0.7% vs. pre 0.2%, p = 0.030). Body-Mass-Index (BMI)-SDS for PWS showed no differences between the groups at any time. BMI-SDS-differences showed no influence on differences in SpO2 evaluations. In the genetic subgroup with deletion there was a statistically significant effect on an increased number of OSA (p = 0.027). The genetic subgroup with uniparental disomy (UPD) was associated with a reduced risk of higher HF (p = 0.035) and less hypopnea (p = 0.011).ConclusionPWS patients predominantly experienced only mild to medium symptoms during COVID-19 infection without necessity of hospitalisation. However, on average three months after infection, differences in PSG evaluations were still apparent, manifesting in lower SpO2 and more frequent hypopnea. A long-lasting impairment of the pulmonary system due to the COVID-19 infection might be responsible.Keywords: Prader-Willi syndrome, COVID-19, Long COVID, Polysomnography, COVID-19 sequelae Introduction Prader-Willi syndrome (PWS) is a genetic disorder affecting 1/10,000 to 1/20,000 live births. It is caused by changes on segment 15q11-q13 of the paternal chromosome and attributed to paternal deletion (60%), maternal uniparental disomy (35%) or imprinting defects [1]. Clinical symptoms change with age, starting with central hypotonia, reduced muscle mass, and feeding problems after birth, which can cause respiratory distress and asphyxia [2]. In early childhood, weight gain occurs around two years of age followed by increased interest in food and then hyperphagia by age five. This leads to weight gain and obesity if energy intake is not controlled [1–3]. The PWS phenotype is characterized by short stature with small hands and feet due to hypothalamic dysfunction [4, 5] and growth hormone treatment [2] is recommended as soon as the diagnosis is made and after adequate nutrition has been achieved, usually between three to four month of age [1, 6]. Because of a high prevalence of sleep disorders and respiratory dysfunction, including oxygen desaturation, central hypoventilation, and obstructive sleep apnoea (OSA), overnight polysomnography (PSG) and the monitoring of oxygen saturation measured by pulse oximetry (SpO2) is recommended annually from childhood onwards [1, 2, 7, 8]. As a reaction to a rapid worldwide spread of novel coronavirus disease 2019 (COVID-19), the World Health Organization declared a global pandemic in March 2020 [9]. COVID-19 clinically manifests itself variably with cough, fever, sore throat, anosmia, and respiratory problems such as dyspnoea [10, 11]. Children are more likely to have mild symptoms or remain asymptomatic during throughout infection [11]. Risk factors for a higher symptom burden are age above 60 years, obesity, male sex, and comorbidities such as OSA, diabetes, hypertension, cardiovascular, and chronic kidney diseases [12, 13]. In paediatric patients obesity is particularly associated with more severe outcomes of a COVID-19 infection [12, 14]. Although patients with PWS seem to be an at-risk group for COVID-19 infection, two prior studies on children and adolescents with PWS showed only a low severity of infection [15, 16]. These studies collected data on the acute course of infection using questionnaires, but did not measure long-term consequences using medical parameters. To our knowledge, there are no studies about possible long-term sequalae of a COVID-19 infection in PWS patients. The aim of this study was to collect data on the acute course of infection as well as on possible long-term sequalae. We studied overnight PSG examinations in PWS patients before and after COVID-19 infection as well as data on the acute course of infection using a questionnaire. Methods Study design and setting In this retrospective and longitudinal study, we collected auxologic and PSG data on 92 children, adolescents, and young adults with PWS who visited a German PWS center for follow-up treatment between April 2022 and November 2022. All members of our cohort had previously participated in the study on effects of COVID-19 lockdown by Mohr et al. in 2021 [17]. Inclusion criteria were two valid PSG examinations and information about the COVID-19 infection status. For detail see Fig. 1 (flow chart). There were 57/92 (64%) participants (35 female) who contracted COVID-19 infection in the time between the first and second PSG. The control group who did not contract COVID-19 consisted of 35 participants (21 female). We compared PSG data from before and after a COVID-19 infection and compared it to the latest two examinations at the PWS center during the same time interval from the control group. Mean time difference between the preliminary and follow-up examination was 6.93 months, 6.85 for the group with the COVID-19 infection and 7.05 for the control group. Flow chart of included and excluded patients, COVID-19 and Control group Mean age was 9.1 with a range from 3.1 to 22 years. There were 78/92 (84.8%) participants treated with growth hormone. The genetic subtype was determined in 72/92 (78.3%) of which 33/72 (45.8%) had a deletion, 28/72 (38.9%) a uniparental disomy, 6/72 (8.3%) an imprinting defect, and 5/72 (6.9%) a non-deletion (not differentiated into uniparental disomy or imprinting defect). In 20/92 (21.7%), the genetic subtype was unknown, as PWS was diagnosed by an analysis of the methylation pattern of chromosome 15q11-13. In the COVID-19 group, the distribution of genetics was as follows: 13/57 (22.8%) deletion, 19/57 (33.3%) uniparental disomy, 2/57 (3,5%) imprinting defect, 3/57 (5.3%) non-deletion, 20/57 (35.1%) diagnosed by analyses of the methylation pattern of chromosome 15q11-13. Further, in the COVID-19 group, 54/57 (94.7%) of parents or guardians gave information on the COVID-19 vaccination status of their children before their COVID-19 infection. There were 32/54 (59.3%) who had not been vaccinated, five (9.3%) once, twelve (22.2%) twice and five (9.3%) three times before their COVID-19 infection. In the control group, 19/35 (54.3%) had not been vaccinated, nine (25.7%) twice (25.7%), six (17.1%) three and one (2.9%) four times against COVID-19. None of the participants received prophylactic steroid treatment for adrenal insufficiency, and none of the participants received treatment with continuous positive air pressure (CPAP) pre, during or post COVID-19 examination. Additionally, in 30 participants who had contracted COVID-19 infection, PSG was not available. These participants had to be excluded from the final sample (n = 30), but we have taken their questionnaires into account concerning evaluation of symptoms (see Fig. 1, flow chart). All procedures performed in this study involving human participants were in accordance with the ethical standards of the University Hospital Bonn and with the 1964 Helsinki Declaration and its later amendments. Ethical approval was given by the institutional Ethics Committee of the University of Bonn, Germany (ethics committee number: 31/15; June 24, 2020). All parents or guardians gave written informed consent for participation. Respiratory analysis For respiratory analysis and SpO2 measurements, a SOMNOtouch™ RESP by SOMNOmedics was used as a PSG device, including paediatric sensors for children. Basal SpO2, lowermost SpO2, number of desaturations, oxygen desaturation index (ODI), amount of sleep time in optimal saturation range (95–100%), suboptimal saturation range (90–95%), and poor saturation range (85–90%) were measured. If a nasal cannula was tolerated, hypopnea, OSA, mixed apnoea, and central apnoea were differentiated. An electrocardiogram measured heart rate. Body composition Height was measured using a standardized rigid stadiometer and weight was measured using a standardized mechanical scale for weight measurement to the nearest 0.1 kg. Body Mass Index (BMI) was calculated as weight/height2 and standard deviation scores (SDS) adapted to age and sex for height, weight, and BMI were calculated. Reference values for children with PWS (SDSPWS) [18] and healthy German children (SD healthy) [19] were used. The questionnaire The participants’ parents or guardians answered our questionnaire retrospectively between November 2022 and January 2023 via e-mail or telephone. The amount of time that had passed since the COVID-19 infection therefore varies accordingly. All COVID-19 infections occurred between January 2020 and November 2022, with only one participant having an infection in 2020, six in 2021, and the majority of our participants having been infected with COVID-19 in 2022. Questions on COVID-19 infections included date of diagnosis, onset of symptoms, type of testing (personal antigen-test, official antigen-test from a certified test centre or PCR) and symptoms experienced. In asymptomatic participants the positive test result was detected by routine tests at school, the residential facility or when taking part in support measures and leisure activities. The symptoms we included into the questionnaire were inspired by German COVID-19 report DEMIS (Robert Koch Institute). We subdivided the COVID-19 group into a mild, medium, and severe COVID-19 infection based on the reported symptoms by the parents or guardians and long-term impact of the COVID-19 infection (see Fig. 2). Low severity was defined as an asymptomatic infection or an infection including symptoms two to ten of Fig. 2 during the COVID-19 infection (see Fig. 2). Medium severity was defined as incomplete remission of all symptoms after infection as reported by caregivers, secondary disease, dyspnoea, tachypnoea, or tachycardia. A severe COVID-19 infection was defined as requiring ventilation, intensive care unit treatment, or death. The parents or guardians were asked to give information on COVID-19 vaccination status and how the vaccination was tolerated by their children. Additionally, data on other infections besides COVID-19 within the last year were collected. Reported symptoms of COVID-19 infection in PWS patients Statistical analyses Statistical analyses were carried out with IBM SPSS Statistics (version 27). Standard deviation scores were calculated for height, weight, and BMI, based on German reference values [19] and PWS reference values [18]. If patients were older than 18 years, SDS values for 18 years were used. Clinical data and results of PSG are reported as means and standard deviations (SD). Delta values (Δs) indicate differences between pre and post COVID-19 follow-up. To calculate Δ, for the BMI-SDS difference, the data from the pre-infection-exam were always subtracted from the data of the post COVID-19 follow-up (post COVID-19 BMI – pre COVID-19 BMI = individual Δs). No significant differences in post-pre-differences (Δs) were found between female and male pairs, making a further division into female and male subgroups superfluous. Before testing, each parameter was always evaluated for normal distribution. If the assumption of normality did not hold or group size was n < 10, non-parametric testing was performed. For measurements on ratio scales e.g., auxological data and results of PSG, a paired t-test or the Wilcoxon signed-rank test were used to compare pre- and post-values. For analyses between COVID-19 and control group, independent sample t-test or Mann-Whitney-U-Test were used. In order to further analyse determinants of differences in sleep laboratory examinations, we performed stepwise multiple linear regression analyses. The individual Δs were always used as dependent variables. Independent variables included. into the analysis were: sex, age, and ΔBMI-SDShealthy. To further analyse the influence of COVID-19 infection, a second model of stepwise multiple linear regression analyses was performed, with COVID-19 symptoms (ordinal, 0 = no symptoms, 1 = mild symptoms, 2 = moderate to severe symptoms) and time since COVID-19 infection in days as independent variables. We decided to perform two separate analyses instead of one approach to avoid overfitting of the model with n = 57 in the COVID-19 group, as advised by Babyak and Rothman (p. 226 of 2,3). To further analyse if the individual Δ (= post COVID-19 – pre COVID-19) was influenced by the genetic subtype, dummy variables were coded for each genetic subtype. Stepwise linear regression was performed for each individual Δ of all parameters with all of the dichotomous variables for genetic subtypes. Cohen’s f² values were calculated to estimate effect sizes. Significance was defined as p ≤ 0.05, a trend as p > 0.05 < 0.1. Since this work represents an exploratory data analysis, no adjustments for multiple testing were performed [20, 21]. Results Severity of illness and symptoms of COVID-19 infection Figure 2 shows the reported symptoms of COVID-19 infection for all affected patients and for those in which repeated PSG was available. According to the reported symptoms we classified the course of the COVID-19 infection as asymptomatic in 8/82 (9.8%), mild in 63/82 (76.8%), and medium in 11/82 (13.4%). There were no severe outcomes in our study cohort. Fatigue (n = 6, 7.3%), exhaustion (n = 5, 6.1%), stomach ache (n = 2, 2.4%), night sweats (n = 2, 2.4%), hair loss (n = 1, 1.2%) and loss of appetite (n = 1, 1.2%) were reported as free text answers in the question on other symptoms. No parents or guardians reported a loss of taste or pneumonia. None of our participants was hospitalised. Only one participant visited the emergency room, because of shortness of breath and received outpatient treatment. In 78/82 (95.1%) there was a total remission of all symptoms reported by caregivers. Three caregivers claimed that their child is still less resilient than before the infection and one participant developed rheumatism of the ankle and hand joint. Inter-group differences at pre-examination There were no statistically significant differences for PSG measured between the COVID-19 and control group at the pre-COVID examination apart from more frequent mixed apnoea in the COVID-19 group (COVID-19 group 2.56 vs. control group 1.00; p = 0.042). Table 1 illustrates this in more detail (see Table 1). Table 1Intra-group comparison between pre/post data in the control group; intra-group comparison pre/post COVID-19 infection in the COVID-19 group and inter- group comparison (COVID-19 vs. Control group) at pre and post examination8.3 years; 3.1–22.09.5 years; 3.2–20.9Intragrouppre vs. post examination(Control group)n = 35Intragrouppre vs. post COVID-19(COVID-19 group)N = 57IntergroupControl vs. COVID-19 group(n = 92)nResultcontrol groupSDpranks- ; + ; =nResultCOVID-19 groupSDpranks- ; + ; =pHF post [bpm]3588.2018.610.02822; 12; 15785.4413.55ns29; 25; 3nsHF pre [bpm]3592.1418.425786.7216.18nsBMI post [kg/m2]3517.373.320.00610; 25; 05719.345.420.00316; 37; 30.086BMI pre [kg/m2]3517.013.205718.785.02nsBMI SDShealthy1 post350.151.03ns12; 23; 0570.561.230.01424; 32; 1nsBMI SDShealthy1 pre350.031.04570.401.25nsBMI SDSPWS2 post35-1.410.50ns19; 15; 157-1.290.73ns29; 28; 0nsBMI SDSPWS2 pre35-1.400.6557-1.290.77nsmean SpO2 post [%]3595.311.680.08320; 11; 45794.821.540.00130; 10; 17nsmean SpO2 pre [%]3595.861.565795.651.40nslowest SpO2 post [%]3587.893.72ns17; 9; 95686.215.160.00330; 12; 14nslowest SpO2 pre [%]3588.832.945687.2910.79nsdesaturations post [no.]3521.9413.46ns18; 16; 15426.6520.37ns20; 31; 3nsdesaturations pre [no.]3524.8322.755429.7834.42nsODI post332.861.80ns17; 14; 2523.442.45ns23; 26; 3nsODI pre333.313.09523.834.02nsOSA post [no.]2612.239.62ns12; 13; 13817.513.25ns15; 22; 10.095OSA pre [no.]2614.6217.353816.5313.87nshypopnoea post [no.]229.556.97ns10; 12; 03513.9113.380.03511; 22; 2nshypopnoea pre [no.]227.686.123510.669.56nsCA post [no.]105.403.10ns2; 8; 06.064.35ns10; 6; 1nsCA pre [no.]104.303.68176.295.27nsMA post [no.]42.000.82ns0; 3; 192.001.32ns3; 2; 4nsMA pre [no.]41.000.0092.561.880.042time 95–100 SpO2 post [%]3568.9734.76ns21; 13; 15554.2936.850.00139; 13; 30.074time 95–100 SpO2 pre [%]3578.2929.305573.7928.49nstime 90–95% SpO2 post [%]3529.5334.99ns14; 20; 15545.4936.780.00113; 39; 30.042Time 90–95% SpO2 pre [%]3520.8728.885525.8128.23nstime < 90% SpO2 post [%]350.270.71ns6; 7; 22550.702.660.0304; 16; 35Time < 90% SpO2 pre [%]350.180.59550.220.80ns1=Standard deviation scores for healthy children, 2=Standard deviation scores for children with Prader-Willi syndrome, ranks: - indicates post < pre, + indicates post > pre, = indicates post = preAbbreviations: HF = Heart rate, BMI = Body Mass Index, SpO2 = Oxygen saturation measured by pulse oximetry, no. = Number of, ODI = Oxygen desaturation index, OSA = Obstructive sleep apnoea, CA = Central sleep apnoea, MA = Mixed sleep apnoea, time 95–100%= Time in optimal saturation range, time 90–95%= Time in suboptimal saturation range, time < 90%= Time in poor saturation range Intra-group comparison between pre/post data in the control group; intra-group comparison pre/post COVID-19 infection in the COVID-19 group and inter- group comparison (COVID-19 vs. Control group) at pre and post examination 1=Standard deviation scores for healthy children, 2=Standard deviation scores for children with Prader-Willi syndrome, ranks: - indicates post < pre, + indicates post > pre, = indicates post = pre Abbreviations: HF = Heart rate, BMI = Body Mass Index, SpO2 = Oxygen saturation measured by pulse oximetry, no. = Number of, ODI = Oxygen desaturation index, OSA = Obstructive sleep apnoea, CA = Central sleep apnoea, MA = Mixed sleep apnoea, time 95–100%= Time in optimal saturation range, time 90–95%= Time in suboptimal saturation range, time < 90%= Time in poor saturation range Intra-group differences between two consecutive examinations (control group) or pre and post COVID-19 infection (COVID-19 group) In the COVID-19 group, we found a statistically significant reduction of sleep time spent in optimal saturation range of 95–100% (post 54.3% vs. pre 73.8%; p = 0.001) (Fig. 3) and an increase in time in 90–95% saturation range (post 45.5% vs. pre 25.8%; p = 0.001) (Fig. 4), For all data, see Table 1. The average time interval between infection and post-COVID-19 evaluation was 96.2 days, with a range from five to 402 days. Sleep time spent in optimal saturation range (SpO2 95–100%) pre (grey) and post (black) COVID-19 infection in 57 PWS patients Sleep time in reduced saturation range (SpO2 90–95%) pre (grey) and post (black) COVID-19 infection in 57 PWS patients Table 2: Stepwise multiple linear regression analyses of Δ sleep laboratory (Δ = post Covid – pre Covid), 1 0 = female, 1 = male, ² 0 = without symptoms, 1 = mild symptoms, 2 = moderate/severe symptoms. Abbreviations: SpO2 = oxygen saturation measured by pulse oximetry, ODI = oxygen desaturation index, OSA = obstructive sleep apnoea, CA = central sleep apnoea, MA = mixed sleep apnoea, HF = heart rate. Stepwise multiple regression analyses of the COVID-19 subgroup To analyse factors influencing the change in the COVID-19 group between post COVID-19 and pre COVID-19 examination, we performed three multiple regression analyses of Δ = post-COVID-19 – pre COVID-19 (Table 2). In the first multiple regression (results with influence see Table 2), age, gender and ΔBMI Kromeyer showed no influence on Δ: mean SpO2, lowest SpO2, time spent in optimal saturation range (95–100% SpO2), time spent in suboptimal saturation range (90–95% SpO2), time spent in poor saturation range (< 90% SpO2), number of hypopnoeas, CA, and MA. Table 2Stepwise multiple linear regression analyses of Δ sleep laboratory (Δ = post COVID-19 – pre COVID-19) 10 = female, 1 = maleStepwise multiple linear regression analyses in COVID-19 subgroupIndependent variables regression standardised coefficients (ß)Dep. variablesSex1AgeΔ BMISDShealthyR2 corr.Δ desaturationsp = 0.0860.303p = 0.0260.075, p = 0.026F = 5.27, f²=0.081Δ ODIp = 0.0630.280p = 0.0450.060, p = 0.045F = 4.24, f²=0.064Δ OSA-0.353p = 0.0300.100, p = 0.030F = 5.12, f²=0.11Abbreviations: BMI SDShealthy= body mass index standard deviation score for healthy German children, ODI = oxygen desaturation index, OSA = obstructive sleep apnoea, HF = Heart rate Stepwise multiple linear regression analyses of Δ sleep laboratory (Δ = post COVID-19 – pre COVID-19) 10 = female, 1 = male 0.303 p = 0.026 0.075, p = 0.026 F = 5.27, f²=0.081 0.280 p = 0.045 0.060, p = 0.045 F = 4.24, f²=0.064 -0.353 p = 0.030 0.100, p = 0.030 F = 5.12, f²=0.11 Abbreviations: BMI SDShealthy= body mass index standard deviation score for healthy German children, ODI = oxygen desaturation index, OSA = obstructive sleep apnoea, HF = Heart rate In a second regression, days passed between COVID-19 infection and PSG evaluation post-COVID-19 infection and symptom severity showed no influence on any of the Δ respiratory parameters analysed. There was no influence of days passed between COVID-19 infection and PSG evaluation post-COVID-19 infection on the Δ of sleep time spent in optimal saturation range of 95–100% (Fig. 5). Regression days since infection and Δ sleep time spent in optimal saturation range 95–100% A third multiple regression analysis showed different effects on Δ HF, ΔOSA and Δ hypopnea in the different genetic subgroups (see Table 3). Table 3Stepwise multiple linear regression analyses of Δ sleep laboratory (Δ = post COVID-19 – pre COVID-19)Stepwise multiple linear regression analyses in COVID-19 subgroupIndependent variables regression standardised coefficients (ß)Dep. variablesDeletionUPDR2 corr.Δ HF-0.280p = 0.0350.062, p = 0.035F = 4.7, n = 57, f²=0.07Δ OSA0.350p = 0.0270.105, p = 0.027,F = 5.33, n = 38, f²=0.12Δ hypopnoeas-0.425p = 0.0110.156, p = 0.011,F = 7.27, n = 35, f²=0.18Abbreviations: ODI = oxygen desaturation index, OSA = obstructive sleep apnoea, HF = Heart rate, UPD = uniparental disomy Stepwise multiple linear regression analyses of Δ sleep laboratory (Δ = post COVID-19 – pre COVID-19) -0.280 p = 0.035 0.062, p = 0.035 F = 4.7, n = 57, f²=0.07 0.350 p = 0.027 0.105, p = 0.027, F = 5.33, n = 38, f²=0.12 -0.425 p = 0.011 0.156, p = 0.011, F = 7.27, n = 35, f²=0.18 Abbreviations: ODI = oxygen desaturation index, OSA = obstructive sleep apnoea, HF = Heart rate, UPD = uniparental disomy Discussion With this study on the acute course of COVID-19 infections in PWS patients and the long-term impact on polysomnographic evaluation, we found subtle but significant and lasting changes detected by PSG. It was hypothesized that patients with PWS would be an at-risk group for a severe course of a COVID-19 infection because of the syndrome associated with obesity and respiratory dysfunction. However, in accordance with others [15, 16], we found only mild to moderate impairment directly associated with the infection. However, through analysis of PSG parameters and oxygen saturation, long-term lasting changes to the respiratory system were detected. Similar to the two mentioned studies [15, 16], respiratory problems and breathing difficulties were only reported twice, tachypnoea occurred four times in our cohort. Fever was the most frequently observed symptom reported by caregivers in our study cohort. Despite the potential for lack of appropriate febrile response, which is often seen in PWS patients due to disturbed thermoregulation [22], we suggest that fever or raised body temperature is a reliable symptom of COVID-19 infection in PWS patients, as it is in healthy individuals [10, 23]. In general, in our study the reported clinical symptoms during the infection observed by caregivers did not differ from those already reported in studies on healthy children [10, 23] and is in accordance with other studies including children and adults with PWS [15, 16]. Thus, we assume that PWS patients have similar symptom profiles in comparison to the general population during the acute infection. Although the severity of infection was only mild to moderate, and the caregivers reported a total remission of the symptoms in almost all participants, there were statistically significant differences in PSG at follow-up examination after an average of three months post-COVID-19 infection. Time of sleep spent in the target area of 95 to 100% oxygen saturation was significantly less in comparison to the examination before COVID-19 infection. The changes in the lowermost SpO2 and mean SpO2 were small, so the clinical relevance of these two changes remains unclear. However, this was clearly a directional trend since the majority of participants showed worsened lowest and mean SpO2 after COVID-19 infection. Moreover, there was a statistically significant shift to more frequent hypopnoea in the post-COVID-19 follow-up examination. This is in accordance with what has already been reported in non-PWS COVID-19 patients. Low oxygen saturation without dyspnoea during the infection and during the acute rehabilitation phase has been described as “happy or silent hypoxia” [24–30]. The reason for the discrepancy between decreased perception of oxygen deficiency and the absence of the feeling of dyspnoea is not fully understood [30]. However, reduced SpO2 during the COVID-19 infection seems to affect PWS patients, as it has been described in a previous study on children and adults with PWS [16]. Whittington et al. reported that “low measured oxygen saturation” was mentioned as a symptom, although it was not explicitly asked for in their questionnaire. They reported that it occurred in “less than 10%” of the patients [16, 31]. It remains unclear, whether reduced SpO2 would have been mentioned more frequently, if it had been studied in routine SpO2 measurements in the acute phase, or if low measured SpO2 would have been added in the dropdown menu of the survey. SpO2 was not measured during the acute infection in our patients, because no patients were hospitalisedduring their infection. PWS patients might be an at-risk group for “happy hypoxia” in COVID-19 infection. In earlier studies, Arens R and Gozal D et al. already showed a non-existent or significantly attenuated hypoxic ventilatory response in patients with PWS that was independent of the degree of obesity [32]. In a further study, they also showed a blunted hypoxic arousal response during sleep in PWS patients compared to a non-PWS control group [33]. They concluded that the primary abnormality of ventilatory control in PWS patients involves the peripheral chemoreceptor pathways [31–33]. Thus, an abnormal ventilation control with insensitivity to O2 and CO2 levels might contribute to low measured SpO2 with absence of dyspnoea in PWS patients with COVID-19 infection. However, because there is still a reduction of SpO2 during sleep after an average of three months’ time after the acute COVID-19 infection, we assume that reduced SpO2 during sleep might be a persisting problem since the acute infection. Further studies are required to prove this hypothesis. Raised BMI and obesity are strong confounders for sleep related breathing disorders (SRBDs) in simply obese children and children with obesity and PWS [34]. Therefore, we wanted to rule out that an increase in BMI might be the reason for the deterioration of SpO2 and more frequent hypopnea during sleep after an average of three months after COVID-19 infection. Mean BMI-SDSPWS remained in the normal range, and the majority of our PWS patients were not obese. Additionally, BMI-SDS for healthy children remained in the normal range, despite a small but significant increase after COVID-19. However, multiple regression analysis of BMI SDShealthy did not show statistically significant influence on the results of the respiratory SpO2 parameters. Thus, a change in BMI SDShealthy could be due to PWS specific phenotypic and metabolic characteristics, as well as changes in nutritional and eating behaviour with increasing age [3, 22]. As these are not typical in children and adolescents without PWS, PWS syndrome specific BMI SDS are calculated [18]. Another possible explanation for our results might be alterations in pulmonary tissue or muscles of respiration. A study on adult non-PWS patients with OSA reported a significant increase in median CPAP pressure in their auto-CPAP treatment after COVID-19 infection [35]. Apnoea-hypopnoea index was slightly increased, but not significantly [35]. So, assuming there was no increased obstruction rate due to possible upper airway changes, e.g. persisting tonsil hyperplasia, this might imply some form of alteration in respiratory mechanics. We found no increase in obstructive sleep apnoea in our PWS patients after COVID-19 infection, but there was an increase in the number of hypopnoeas. More frequent shallower breathing due to alterations in respiratory mechanics might lead to a reduction of SpO2 measurement. This could be caused by changes in pulmonary tissue or muscles of respiration, or both. To date, there have only been few studies in other paediatric patient populations comparing respiratory parameters in the same patients before and after COVID-19 infection and then comparing them with a non-COVID-19 control group. Mogensen et al. compared spirometric lung function tests in a population-based sample of young, healthy adults with and without asthma and a mild-to-moderate COVID-19 disease before COVID-19 infection with spirometric lung function tests after COVID-19 infection [36]. They found no evidence that an infection with COVID-19 resulted in impaired spirometric lung function [36]. In contrast, in a study by Soyak Aytekin et al. mean expiratory flow 25–75% values were significantly reduced after COVID-19 infection in some paediatric patients with asthma compared to prior lung function tests [37]. Thus, the low SpO2 values during sleep in our cohort might also be caused by a small airway dysfunction after COVID-19 [37]. Furthermore, a disturbed gas exchange after COVID-19 has already been discussed as a possible consequence of COVID-19 infection. Wu et al. reported in their study on 3-, 6-, 9-, and 12 months respiratory outcomes in adults without PWS who were hospitalized with severe COVID-19 infection, that diffusion capacity of the lungs for carbon monoxide was 77% of predicted at three months, 76% of predicted at six months, and increased to 88% of predicted at twelve months after hospital discharge [25]. The forced vital capacity also increased from 92% of predicted capacity after three months to 98% of predicted after twelve months [25]. As the diffusion capacity of the lungs for carbon monoxide improved with time [25], the question arises, whether the reduction in SpO2 during sleep in our study might be a consequence of a disturbed gas exchange and therefore, an improvement in SpO2 could also be expected with time. So far, our data does not indicate an improvement with time, as time since infection did not show an influence on PSG parameters in multiple regression analysis. As our study results are based on an average time span of three months after COVID-19 infection, further long-term studies are needed to better assess the time frame in which lowered SpO2 at sleep persists and to evaluate the potential resulting clinical implications of persisting increased sleep time in the lowered SpO2 range of 90–95% SpO2 after COVID-19 infection. For example, in a study by Lau et al.., an impact on verbal memory performance was associated with SpO2 nadir during sleep in children with OSA [38]. In their study, the group without OSA and better verbal memory performance showed a mean SpO2 nadir of 94.82% (range: 89–98) vs. 91.52 (range 76–97) in the OSA group with lower performance. The mean nadir in both groups was above 90%, once 92% and once 95%. So, it could be assumed that not only mean SpO2 < 90% might have clinical consequences, but also differences in the range of 90–100% with relatively more sleep time spent in lower intervals of 90–100%. Another possible clinical implication of reduced SpO2 during sleep is potential neurocognitive deficits, which are associated with sleep disordered breathing in children with sleep related breathing disorders [39]. Further studies are clearly needed to gain more data on clinical implications. Regular follow-up examinations by means of neurocognitive testing in combination with more frequent PSG are important to identify clinical implications after COVID-19 infection and should be considered for the regular care after COVID-19 infection. Since cognitive disability and speech and language delay are known clinical manifestations in PWS [22], particular attention should be paid to the development of these parameters after COVID-19 infection, and possible treatment options should be discussed in time. CPAP, which is already used for PWS patients in the setting of OSA [40], could be considered as a possible treatment option in the case of persisting SpO2 below the optimal range during sleep after COVID-19 infection in PWS patients. According to a study by Turner et al., CPAP treatment had significant positive effects on working memory, long-term verbal memory, and short-term visuospatial memory in patients with OSA [41]. One limitation to our study is the possibility that some members of our control group did in fact contract COVID-19 during this time but remained asymptomatic and as such were not tested. Because of the retrospective study design, some symptoms the PWS patients experienced during their infection might not have been mentioned by the caregivers because of recall bias or because they have not been observed, due to a high pain threshold, which is a typical feature in people with PWS [1, 22]. The young age of our study cohort may have also contributed to a milder course of the disease [42–44]. Because of the low proportion of PWS patients with three or more vaccinations before their COVID-19 infection, no statistical analyses investigating the influence of vaccination status on course of COVID-19 infection and PSG parameters were performed. Since we do not have a non-PWS control group, we cannot comment on whether the effect of COVID-19 infection on PSG parameters noted in this study apply to the general population. Whether a transfer of our results is possible should be investigated in future research. Moreover, we found only mild differences and the clinical implications remain unclear. However, all of our results are unidirectional and still apparent after approximately three months after COVID-19 infection, even if the majority of parents reported no persistent clinical overt symptoms. Further studies are needed to determine clinical significance and duration of the deterioration of PSG in patients with PWS. The comparison between COVID-19 and the control group at post-examination showed a tendency towards more OSA in the COVID-19 group compared to the control group, but when analysing the COVID-19 subgroup, there were no statistically significant changes in HF and the number of OSA and hypopnoeas between pre- and post- COVID-19 examination. However, statistically significant differences were found when the genetic groups in the COVID-19 group were analysed separately. We observed that the deletion genetic subgroup had a statistically significant effect on an increased number of OSA, whereas the UPD genetic subgroup was associated with a reduced risk of higher HF and fewer hypopnoeas. There only have been few studies on the influence of genetics on cardiovascular and pulmonary risk in PWS patients so far. In an earlier study by Torrado et al.., more frequently present sleep disturbances were described in PWS patients with a deletion [45]. Moreover, a recent study by Cintra, Rocha et al. showed more frequent OSA in the deletion genetic group compared to the UPD group and 0% of the patients with an UDP had hypoventilation [46]. This is in accordance with the findings in our study and suggests that the genetic deletion may have a higher risk over time for sleep-disordered breathing and especially for OSA, whereas UPD might be associated with better progression over time. Future studies, with focus on the effect of genetics on cardiopulmonary parameters in a larger study cohort are of importance to prove this hypothesis. The strengths of our study include the large patient population for PWS with COVID-19 infection, the large patient population without COVID-19 infection, and the assessment of longitudinal data. This allows us not only to compare changes pre- and post- COVID-19 for each participant individually, but also to compare the collectives, with and without COVID-19 infection. By regular follow-up examinations due to PWS, we were also able to exclude participants from whom only one valid examination was available or who were lost to follow-up. To the best of our knowledge, there are no prior studies in other paediatric study populations that directly compare longitudinal individual PSG evaluations before and after a COVID-19 infection. Although on 4th May 2023 the World Health Organization declared COVID-19 to be an ongoing and established health issue which no longer constitutes a public health emergency of international concern [47], our study shows that we should continue to conduct studies on COVID-19 infection in PWS, in order to learn more about possible long-term consequences and possible treatment options to ensure optimal patient care. Conclusion In our study cohort, PWS patients experienced only mild to medium symptoms during a COVID-19 infection and no participant was hospitalised. Nevertheless, after an average of three months following COVID-19 infection, differences in PSG were still apparent when compared to pre-Covid-19 examinations, manifesting in lower oxygen saturations and more frequent hypopnoea. In comparison, we observed no differences in two consecutive examinations in our control group. It remains unclear what mechanism caused this deterioration in the COVID-19 group and how long it will last. More studies are needed to evaluate possible long-term consequences and the resulting clinical implications in treatment and frequency of follow-up PSG evaluations for PWS patients. Long-term sequalae in non-PWS patients should be considered. Acknowledgements Parts of this study were carried out at Paediatric Centre, St. Bernward Hospital, Hildesheim, Germany. We thank all participants and their families responding to our questionnaire for this study. Abbreviations Prader-Willi syndrome coronavirus disease 2019 overnight polysomnography obstructive sleep apnoea oxygen saturation measured by pulse oximetry continuous positive air pressure oxygen desaturation index standard deviation scores for children with Prader-Willi syndrome standard deviation score for healthy German children body mass index uniparental disomy standard deviations delta values (post COVID-19 – pre COVID-19) Author contributions SB was involved in the data curation, formal analysis, and investigation of the study. SB drafted the initial manuscript. SS contributed to data curation, formal analysis, methodology, and to reviewing and editing the original manuscript. CL contributed to conceptualization, study design, investigation of the study cohort, supervised data acquisition and quality management, and reviewed the manuscript. BG was responsible for the design and the conceptualization of the study. BG oversighted and supervised the study and reviewed the manuscript. Funding Only internal funding. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Open Access funding enabled and organized by Projekt DEAL. Data availability The data that support the findings of this study are not openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request. Data are located in controlled access storage at Children`s Hospital, University of Bonn. Declarations Ethics approval and consent to participate All procedures performed in this study involving human participants were in accordance with the ethical standards of the University Hospital Bonn and with the 1964 Helsinki Declaration and its later amendments. Ethical approval was given by the institutional Ethics Committee of the University of Bonn, Germany (ethics committee number: 31/15; June 24, 2020). Consent for publication All parents or guardians gave written informed consent for participation. Competing interests The authors declare that they have no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Sandra Schulte and Bettina Gohlke contributed equally to this work. References Associated Data Data Availability Statement The data that support the findings of this study are not openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request. Data are located in controlled access storage at Children`s Hospital, University of Bonn.