This cross-sectional online survey was conducted from January to March 2018. Participants were recruited via the patient database of the NRCHD, the largest European registry for CHD patients (21). For patient recruitment, the database was searched for patients aged between 6 and 17 years at the time of the survey. Respective individuals and their parents were contacted and invited to take part in the survey via email. Ethical approval was obtained by the institutional ethical committee. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki.

PA and sports participation were assessed using the validated comprehensive questionnaire of the “Motorik-Modul” (MoMo) from the German Health Interview and Examination Survey for Children and Adolescents (KiGGS), thus allowing the comparison of obtained data with a representative age-matched subset of 3.385 same-aged participants of the MoMo wave 2 study (2015–2017) (22). Design and results of the MoMo Baseline and Longitudinal Study and details on the structure and content of the MoMo Physical Activity Questionnaire (MoMo-PAQ) have been published previously (22–25). Briefly, it consists of 28 items regarding frequency, duration and intensity of PA to capture habitual PA in different domains (PA in sports clubs, leisure time PA outside of sports clubs, extra-curricular PA, outdoor play, active commuting to school). Furthermore, the MoMo-PAQ consists of 36 items, which assess physical self-description based on the German version of the Physical Self-Description Questionnaire with answer categories to a 4-point Likert Scale (26). These 36 items represent the basic functions of physical performance: strength, endurance, speed, skills, coordination, and flexibility.

In addition, MoMo-PAQ included the German version of the Physical Activity Enjoyment Scale for children and adolescents (PACES) (27). The scale consists of 16 items (nine positive poled items, seven negative poled items) with answer categories to a 5-point Likert Scale.

The questionnaire for CHD-patients was amended by eight specific questions capturing the medical background of sports recommendations, sports restrictions and access to sports.

Values of continuous variables are reported as mean ± standard deviation. The Pearson's chi-square test was used for group comparisons including nominal data (e.g., gender and age). In order to assess the impact of potential contributing factors on PA, analysis of variance and analysis of covariance, Pearson's correlation as well as multiple and linear regression analysis was used, as appropriate. IBM SPSS statistics version 25.0 (IBM Inc., Armonk, NY, USA) was used for statistical analyses. A significance level of p ≤ 0.05 was applied.

Of 21.354 eligible patients, invitation was successfully delivered to 14.496 patients. 1.718 patients decided to participate in the study. Only complete datasets were considered for evaluation and were available from 1.198 CHD patients with a mean age of 11.6 ± 3.1 years. Of those 53.8% were male, and 46.2% female. The study participants were allocated into simple, moderate and complex CHD classification according to Warnes et al. (28) (Table 1). Thus, 411 (34.3%) were classified as simple CHD, 423 (35.3%) moderate, and 364 (30.4%) as complex CHD. Included patients had untreated CHD in 49.3%, 1–3 operations/interventions in 30.2%, and more than three operations/interventions in 20.5%. Genetic syndromes and chromosomal disorders were present in 70 patients (5.8%), most frequently trisomy 21 in 37 patients (3.1%) and Di-George-Syndrome in 15 patients (1.3%). Of all patients, 57.2% stated to live in an urban or suburban environment, whereas 42.8% live in rural areas. Patient characteristics of the study group did not significantly differ to the entire cohort of eligible patients.

Compared to MoMo participants, CHD patients reached significantly less frequently the WHO recommended level of 60 min of daily PA (8.8 vs. 12%; p < 0.001), simple CHD 9.2%, moderate 9.2%, and complex CHD 8.0%. Children with CHD were 0.62 days per week less active than those of the reference group (p < 0.001) (Figure 1).

Participation in sports clubs was significantly reduced in children with complex CHD compared to MoMo participants (53.6 vs. 67.6%; p < 0.001), whereas patients with simple and moderate CHD showed similar frequent participation as the reference group (74.9 and 66.2%, respectively). 61.4% of children with simple, 56.8% with moderate and 47.7% with complex CHD stated to participate even in competitive sports, compared to 53.1% of the reference group.

Enjoyment in sports was almost equally distributed across CHD groups and MoMo participants. Using the positive dimension of the PACES Scale ranging from 9 to 45 points we found 35.9 points for patients with simple CHD, 34.6 points for moderate and 33.2 points for patients with complex CHD, compared to 35.2 points for the reference group. Enjoyment in sports correlated with the level of PA (r = 0.41; p < 0.001).

Physical self-description of all basic functions of physical performance (strength, endurance, speed, skills, coordination, flexibility) appeared to be related to the level of PA (Figure 2). More physically active patients showed higher percentage of positive self-description of basic functions of physical performance, most consistent for the dimension “endurance.”

Further analyzing the physical self-description of the dimension “endurance” (ranging from 6 to 24 points) showed significant reduction in CHD patients compared to the reference cohort (p < 0.001), most noticeable in complex CHD (11.9 ± 4.9 points) compared to 16.6 ± 4.5 points in the reference group (Figure 3). The dimension “endurance” correlated with the level of PA in all CHD groups (p < 0.001).

49.2% of the children with complex CHD, 31.7% with moderate and even 13.1% with simple CHD were advised by their physician to restrict PA. 2.3% of all CHD patients received even a complete dispensation. Most frequently, sports restrictions have been advised in patients with Ebstein's anomaly, followed by single ventricle hemodynamics, and aortic valve stenosis (Figure 4). Interestingly, also patients with more simple CHD (i.e., ventricular septal defect or atrial septal defect) received sports restrictions. Advisory physicians in terms of sports were predominantly pediatric cardiologists (55.4%), followed by pediatricians (24.9%), family physicians (9.2%), orthopedists (2%), and sports medicine specialists (1.1%). In 7.4% the participant was not aware of the exact professional title of the advisory physician.

Regression analysis demonstrated a significant impact of age and gender on PA, whereas girls with CHD demonstrated lower levels of PA. Both, number of interventions and enjoyment in sports play a relevant role on the level of sports activity. Multiple regression analysis revealed physician-recommended PA restrictions having significant impact on the amount of PA, followed by sports activity of the father, whereas sports activity of the mother and residence in rural areas obviously had no significant impact on patient's PA (Table 2).

Data of this survey are based on self- or proxy-reports and might therefore be prone to bias, including recall bias and social desirability, as children with CHD may overestimate their PA level (34). Furthermore, the absence of direct contact with patients may leave the chance for misunderstanding of questions or incorrect answers. To prevent such occurrences each study participant was offered to contact the study management if having any questions. In addition, at the end of the survey there was the possibility to provide ambiguities and questions directly as free text. Clearly, measuring PA by objective methods (i.e., accelerometry, pedometers, activity trackers) might be more precise (14). However, the objective assessment of PA in different domains can be challenging and requires complex strategies, which can hardly be applied in large patient populations. The strength of the current study is the consideration of different domains of PA, the large sample size and the inclusion of school-aged children with a wide age range, as well as the representative reference group.

Since this is a cross-sectional study, we provide descriptive information and report on associations rather than claiming to report causal relationships between PA reduction and potential contributing factors (2). Moreover, the results reflect respondents' subjective statements. The results may not be applicable to CHD patients outside Germany, since they are affected by the life situation of the patients, as well as the organization of the health care and educational system (2).

Since the patients have been invited by emails to participate in the survey, the response rate of 11.9% was relatively low. Nevertheless, the sample of patients participating in the online survey did not differ significantly to the entire CHD cohort of NRCHD and therefore seems to be representative for the German CHD community at large.

Theoretically, by considering the prevalence of CHD in the German population of 1.1%, it cannot be ruled out that a minority of participants in the KiGGS study may also be affected by CHD. Due to inaccurate recording of CHD diagnoses in the KiGGS dataset a removal of these participants was not feasible. However, from a statistical point of view, having these children included in the reference group should not result in a relevant bias.