IntroductionNewborn screening (NBS) is a key element of secondary prevention in public health [1]. Traditionally, NBS has primarily used biochemical assays to detect conditions meeting specific criteria, such as early onset and availability of effective treatment. In Germany, 23 early-onset conditions (effective May 2026) are screened [2]. Advances in genome sequencing and decreasing costs have increased interest in incorporating genomic sequencing into NBS (genomic newborn screening, gNBS) to identify a broader spectrum of genetic disorders at birth and initiate treatment if available [3,4,5]. A central challenge is deciding which conditions to include [5, 6], especially late-onset, untreatable or uncertain conditions.Parental preferences for gNBS have been explored through qualitative studies [7,8,9,10], questionnaires [11,12,13,14,15,16], vignette-based studies [17, 18], and discrete choice experiments (DCEs) [19, 20], consistently showing high general interest in gNBS. A DCE in Australia found broad public support for gNBS, with screening cost emerging as the most influential attribute [19]. Participants preferred programs with more diagnoses, but acceptance decreased for conditions with limited treatability and uncertain penetrance. In the NC NEXUS DCE on medically non-actionable conditions, penetrance and severity influenced parental choices, whereas the availability of quality-of-life interventions did not [20]. Higher education, health literacy, genetic testing experience, and future child planning increased gNBS interest in some studies [14, 17], while infant-health concerns or conservative views were associated with less support [17]. Gender differences were minor [20], and sociodemographic factors, including family history, showed inconsistent effects [14, 16]. Intentions to pursue gNBS were stronger when participants held positive beliefs about sequencing and completed a values clarification exercise [15].Despite this evidence, several questions remain, particularly in the German context: how expecting parents would evaluate gNBS results differing in actionability and penetrance, which psychological and social factors could shape interest in such results, and how parents might perceive risk after result disclosure. Addressing these questions is important for ethically and socially responsive criteria for the potential implementation in Germany. This study was conducted within the project “NEW_LIVES” (https://www.klinikum.uni-heidelberg.de/en/new-lives-genomic-newborn-screening-programs), which aimed to develop ethical, legal, and social criteria for potential implementation of gNBS in Germany. In this study, we pursued three main aims. First, we examined parents’ general attitudes, preferences (e.g., regarding disclosure criteria and data storage), and emotional responses regarding hypothetical gNBS results, and explored preference profiles. Second, using hypothetical gNBS results that varied in actionability and penetrance, we investigated parents’ interest in receiving such results and examine both condition-related and individual predictors of interest. Actionability and penetrance were selected because previous research has linked limited actionability and lower or uncertain penetrance to parental ambivalence [7]. Third, we assessed parents’ subjective risk perception following hypothetical gNBS results, operationalized as diagnostic concern and subjective probability estimates, examined its correspondence with objective risk (condition penetrance), and identified predictors beyond objective risk.Hypotheses and variable selection were guided by prior empirical evidence and Fuzzy Trace Theory (FTT) [21]. Previous evidence suggests that parents weigh perceived benefits and harms, consider whether testing aligns with their personal values and beliefs, and are influenced by emotional states and individual circumstances, such as family history, when deciding about genetic testing in children [22]. Thus, decisions are not based solely on cognitive evaluations of medical information, but also on emotional and social contexts. This perspective is consistent with FTT, which proposes that people often rely on the “gist” of information, particularly under uncertainty or heightened emotion [21]. Additionally, variable selection was informed by a prior focus group study in Germany [7], in which participants identified social expectations and trends as important influences on gNBS attitudes.For interest in receiving gNBS findings, we distinguished condition-level characteristics from individual-level psychological and social factors. At the condition level, we hypothesized that interest in gNBS would be higher for conditions with greater actionability and higher penetrance [7]. At the individual level, we expected associations with diagnostic concern (worry about the disease occurring in the future) and perceived probability of disease onset. Consistent with FTT [21], we emphasized subjective risk perception because subjective interpretations of risk may be particularly relevant under uncertainty. We also expected interest to be associated with affective, experiential, social, and belief-related factors. Based on research on parental genetic testing decisions, we included attitudes and emotions toward gNBS, chronic condition experience, uncertainty tolerance, and anticipated coping with gNBS results as indicators of how individuals evaluate, respond to, and expect to manage genetic information [23]. We additionally included perceived social norms, as social expectations and trends were identified as potentially influential [7, 24]. Genetic determinism beliefs were included because viewing genes as strong determinants of health may increase the perceived relevance of genetic information. Accordingly, we expected higher interest with more positive attitudes and emotions toward gNBS, chronic condition experience, greater uncertainty tolerance, more positive anticipated coping, stronger perceived social norms, and stronger genetic determinism beliefs.For study aim 3, we examined risk perception in more detail through perceived probability of disease onset and diagnostic concern. Previous studies linked higher perceived risk to family history [23, 25], distress, anxiety, depression, beliefs, and preventability or treatability [23, 26, 27]. While many studies assess risk perception before disclosure, less is known about risk perception after result disclosure. By assessing risk perception after hypothetical result disclosure, we examined responses to gNBS results differing in actionability and penetrance and whether subjective risk corresponded to objective risk, indicated by condition penetrance. We further explored associations with chronic condition experience, gNBS attitudes, actionability, emotions, stress, anxiety, and numeracy disease-onset uncertainty tolerance.Materials and methodsStudy designWe conducted a cross-sectional online survey (06-11/2024) on parental gNBS preferences and attitudes, using a single-choice-experiment design. Participants were exposed to randomized case scenarios and related questions, enabling between- and within-subject comparisons. They were masked for the randomization purposes and scenario variations. The study was approved by the ethics committee of Heidelberg Medical Faculty (S-215/2024) and pre-registered (https://doi.org/10.17605/OSF.IO/APZ6F).ParticipantsEligible participants were expecting parents and/or parents with at least one child under eight years, based on the NEW_LIVES consortium’s recommendation of a median age of disease-onset of seven years as a threshold for gNBS conditions in Germany [28, 29], making parents with (or expecting) children in this age range were considered particularly relevant. Participants had to be 18 years of age or older and proficient in German. Recruitment used flyers, email lists, parent cafés, forums, and patient organizations for families with chronic diseases. Compensation was 10 euros. The target sample size was 540, with ≥30 participants per group for stable estimates.ProcedureParticipants accessed the study online on SoSciSurvey, consented, and answered sociodemographic questions. After information on the current NBS and potential gNBS in Germany, they completed comprehension checks. They rated their agreement with statements reflecting potential advantages/hopes and disadvantages/worries of gNBS, emotions related to gNBS, and criteria for target diseases. Questionnaire-based randomization without replacement assigned each participant to two distinct scenarios from nine, describing gNBS results with varying actionability and genetic penetrance. After each scenario, participants rated emotions, interest in findings, risk perception, and anticipated coping. Additional variables followed. The survey included novel items and validated instruments; Appendix 1 contains the full questionnaire.MeasuresBuilding on a systematic review [22], we selected variables that were (a) consistently associated with testing behavior or interest, (b) particularly relevant for population-wide NBS (as opposed to diagnostic testing), and (c) measurable through self-report surveys.1)Sociodemographics: Age, gender, highest level of education and experience with chronic disease in the close environment.2)Attitudes toward gNBS: Agreement with potential hopes/benefits and concerns/worries regarding gNBS (5-point Likert scale + “don’t know” response option). Items partly adapted from prior gNBS [15] and preimplantation genetic diagnosis (PGD) studies [30], and developed within the NEW_LIVES consortium.3)Emotions toward gNBS: Presence and strength of emotions (0-100 slider), selected based on focus group findings [7] and PGD research [30].4)General gNBS preferences: Age of onset (categorices), penetrance (0-100% slider, for actionable/monitorable/non-actionable conditions), actionability and severity (0-100 sliders), and family-related information (categories). Items developed through discussions (NEW_LIVES consortium) and focus groups [7].5)Case-scenarios: Interest in gNBS, anticipated coping with gNBS results, diagnostic concern (worry about disease manifestation), perceived probability of disease-onset, emotions associated with hypothetical gNBS finding (0-100 sliders). Scenarios varied by actionability (1: no specific treatment, but supportive therapies may help manage symptoms; 2: no effective treatment preventing symptoms and ensuring normal development, but options available that alleviate symptoms and improve quality of life; 3: effective treatment preventing severe symptoms and enabling normal development), and penetrance (1 = 25–49%, 2 = 50-74%, 3 = 75-100%).6)Additional measures: Perceived social norms (3 items, adapted from Link et al. [31] and Paquin et al. [32]), “Tolerance for Ambiguity Scale” [33], item based on the English “Health Literacy” scale [24], anxiety (GAD-2 scale, short form of GAD-7) [34], perceived stress (PSS4-German) [35], numeracy [36], genetic determinism item (adapted from BabySeq Project [37]).Data analysisFor aim 1 (exploring gNBS-related attitudes, emotions, and preferences towards gNBS), general, non-scenario-specific variables were analyzed descriptively. Reliability coefficients assessed internal consistency for newly developed item sets. Responses to positive/negative attitudes and emotion items were aggregated as means and compared with paired-sample t-tests. Group differences in attitude ratings by chronic condition experience and gender were examined using independent-samples t-tests or Wilcoxon rank-sum tests, depending on distributional assumptions. A latent profile analysis (LPA) of actionability, penetrance, and severity preferences identified profiles; post hoc tests compared participant characteristics.For aims 2 and 3—examining interest in and risk perception of abnormal gNBS findings and their predictors—responses to the hypothetical scenarios were analyzed using multilevel models (MLMs; lme4 package [38]), accounting for the nesting of scenarios within participants. Interest in hypothetical gNBS results was modeled as an outcome for aim 2; for aim 3, perceived probability of disease onset and diagnostic concern were modeled separately to identify predictors of subjective risk perception. Full maximum likelihood estimation was applied. Predictors were centered before analysis (level 1: person mean; level 2: grand mean) [39]. Model assumptions were checked using residual plots and the DHARMa package [40]. Due to heteroscedasticity, outcomes were transformed (interest: reflection and Box-Cox transformation; perceived probability and diagnostic concern: logit transformation), and robust standard errors were computed (clubSandwich package [41]). Sensitivity analyses excluded influential outliers to ensure robustness. Sociodemographic variables not mentioned in the literature were included if marginally significant in bivariate analyses with the outcome variable.Significance level was set at α = 0.05. P values were adjusted for multiple testing using Benjamini-Hochberg correction.ResultsSampleOf 797 survey starters, 13 stopped at the study information, two lacked sufficient language proficiency, and six failed the inclusion criteria (pregnant (partner) and/or child