Abstract
Newborn screening (NBS) is one of the most effective measures of secondary prevention. While the benefit of NBS on the clinical long‐term outcomes of children with inherited metabolic diseases (IMD) has been demonstrated, the potential burden of families living with an early diagnosed and treated child with an IMD has not been thoroughly investigated. The aim of this longitudinal questionnaire‐based study on 369 families living with a child with an IMD was to investigate the psychosocial and financial burden following a true‐positive NBS. The reported psychosocial burden differed between children and their parents, and was associated with the child's age, diagnosis, and treatment. At younger ages, parent‐reported burden was higher for the parents than for the individual child, while it increased for children and decreased for parents as the child grew older. Furthermore, psychosocial burden increased if the child required a strict dietary treatment and was at risk of metabolic decompensation. Regardless of diagnosis and treatment, the developmental delay of their child independently increased the parental psychosocial burden. Financial burden was reported by 24% of all families, and was higher in low‐income families and in families whose children required dietary treatment. In conclusion, a substantial psychosocial and financial burden was revealed for children and their families after true‐positive NBS. Since this burden is likely to have a negative impact on the long‐term individual health benefits of NBS, this study underlines the importance of regularly assessing the psychosocial and financial needs of these families.
Keywords: financial burden, inherited metabolic disease, neonatal screening, newborn screening, psychosocial burden, psychosocial stress
1. INTRODUCTION
Newborn screening (NBS) programs have been developed over the course of more than 50 years along with a continuous expansion of NBS disease panels through the implementation of high‐throughput technologies such as tandem mass spectrometry (MS/MS). 1 , 2 The current German national NBS panel includes 13 inherited metabolic diseases (IMD). 3 The intended beneficial impact on the clinical long‐term outcome including reduction of morbidity and mortality could be proven for single metabolic disorders 4 , 5 , 6 , 7 , 8 as well as for entire screening panels. 9 , 10 , 11 , 12 , 13 , 14 However, it should not be under‐estimated that this success is often considerably offset by the substantial impact of diagnosis and therapy on the life of the affected child and her/his family. Receiving the diagnosis of a chronic disease composes a psychosocial risk 15 , 16 which is especially true for parents of healthy newborns receiving an unexpected pathological screening result unraveling an intrinsic and often severe health risk. 17 , 18 With the possible risk of metabolic decompensation and uncertainty about the child's future development, strict disease management by parents becomes crucial. Additionally, high hospitalization rates and regular outpatient visits are common among IMD patients disturbing family life. 7 , 19 , 20 Accordingly, a significant parental burden has been reported for various metabolic diseases. 15 , 17 , 21 , 22 , 23 , 24 , 25 , 26 , 27 However, the social impact of living with an IMD identified by NBS has been less explored, 21 , 24 , 28 and especially the longitudinal trajectory of these families over the years has been scarcely investigated. In addition to the psychosocial burden, a financial burden might also be imposed on parents if medications and essential dietary products were not fully paid for by health insurances. 17 , 22 , 24 , 25 The aim of this questionnaire‐based multicenter longitudinal study is to investigate in further detail the perceived psychosocial and financial burdens of families with a child affected by an IMD identified by NBS. The evaluation focuses in particular on the long‐term development of the perceived burdens over an observation period of up to 16 years for each individual.
2. PATIENTS AND METHODS
2.1. Study design
To investigate the clinical benefit of NBS for individuals with an IMD identified by NBS, a prospective multicenter longitudinal outcome study was implemented in 2005 (NGS2020/2025, German Clinical Trials Register identifier: DRKS00013329; PI S. Kölker, https://www.klinikum.uni‐heidelberg.de/fachliche‐zentren/dietmar‐hopp‐stoffwechselzentrum/neugeborenenscreening/pilotprojekt‐ngs‐2020, 13 ). The study was approved by the local ethics committee of the coordinating site (Medical Faculty of Heidelberg, application S‐104/2005) and consecutively by all study sites (Düsseldorf, Freiburg, Mainz, Reutlingen, and Ulm, all in Germany). As previously described, 13 inclusion criteria are: (A) date of birth at or after January 1, 1999; (B) confirmation of the screened IMD according to the national guideline 29 ; and (C) written informed consent by the child's caregivers before enrollment. Children who did not fulfill all the above listed inclusion criteria were excluded from the study. Individuals enrolled in the observational study were followed according to a standardized protocol. Comprehensive regular follow‐up information was obtained at study visits at the ages of 1.5, 3.5, 5.5 (all ±0.5), 9, 14, and 18 (all ±1) years by structured clinical examination, recording of medical history, age‐adapted neuropsychological testing and completion of a burden questionnaire. For this evaluation, delayed development and/or abnormal cognition were defined as non‐age appropriate Denver Developmental Screening Test (DDST) or IQ below 85 in age‐adapted Wechsler tests, i.e., Wechsler Preschool and Primary Scale of Intelligence (WPPSI‐III, WPPSI‐IV) and Wechsler Intelligence Scale for Children (WISC‐IV, WISC‐V). Whenever the time elapsed between the last test result and filling out the questionnaire exceeded 1 year, the child's development was rated by expert opinion based on other neuropsychological tests and information about school attendance.
The questionnaire‐based study includes patients affected by one the 13 IMDs of the national NBS disease panel in Germany 30 : Phenylketonuria (PKU; OMIM #261600), tyrosinemia type 1 (Tyr1; OMIM #276700), maple syrup urine disease (MSUD; OMIM #248600), isovaleric aciduria (IVA; OMIM #243500), glutaric aciduria type 1 (GA1; OMIM #231670), very‐long chain acyl‐CoA dehydrogenase deficiency (VLCADD; OMIM #201475), long chain 3‐hydroxyacyl‐CoA dehydrogenase deficiency (LCHADD; OMIM #609016)/mitochondrial trifunctional protein deficiency (mTFPD; OMIM #609015), medium chain acyl‐CoA dehydrogenase deficiency (MCADD; OMIM #607008), carnitine acylcarnitine translocase deficiency (CACTD; OMIM #212138), carnitine palmitoyltransferase 1 (CPT1D; OMIM #255120) and 2 deficiency (CPT2D; OMIM #600650), biotinidase deficiency (BioD; OMIM #253260), and classic galactosemia (Gal; OMIM #230400). Since stratified management depends on the disease severity in IVA and VLCADD, both were subclassified as “attenuated” or “classic” as published before. 7 , 31 In brief, children with IVA with the common IVD gene variant (c.941C>T) and/or an isovalerylcarnitine (C5) concentration in NBS below 6 μmoL/L were classified as attenuated IVA (aIVA) and those with an initial C5 concentration of at least 6 μmoL/L and/or early onset metabolic decompensation as classic IVA (cIVA). 7 Children with VLCADD with an initial tetradecenoylcarnitine (C14:1) concentration up to 1 μmol/L and residual enzyme activity of at least 10% were classified as attenuated VLCADD (aVLCADD), while those with C14:1 above 1 μmol/L, residual enzymatic activity below 10% or elevated serum creatine kinase concentrations were classified as classic VLCADD (cVLCADD). 31 A risk of metabolic decompensation was considered in all children with an IMD requiring the prevention of prolonged fasting and catabolism, specifically in individuals with MCADD, VLCADD, LCHADD/mTFPD, CPT1D, CPT2D, CACTD, IVA, GA1, and MSUD.
2.2. Questionnaire
To evaluate the psychosocial and financial burden, a questionnaire (Supplementary Figure S1) is included in the above described prospective multicenter longitudinal outcome study and is completed by the proband's parents at each study visit at the ages of 1.5, 3.5, 5.5 (all ±0.5), 9, 14, and 18 (all ±1) years. Endpoints included in this questionnaire (Supplementary Figure S1) are the following: parents' subjective perception of (1) psychosocial burden for the child and her/his parents rated on a 5‐point Likert scale (no, mild, medium, severe, and very severe burden); (2) financial burden; (3) negative impact of the metabolic disorder on mother's or father's professional lives; and (4) expectation of an independent life for their child. In addition, the following co‐variates are included in the questionnaire (Supplementary Figure S1): (A) Net family income, classified to three categories following data of the German Federal Statistical Office 32 and adapted due to inflation during the conduct of the study, lastly: low income: representing average for unskilled positions (<2000 € net family income per month); medium income: representing average for skilled positions (2000–3500 €); high income: representing average for managerial positions (> 3500 €). (B) Additional familial stress factors that may impact the psychosocial burden, e.g., other family members in need of care, with disabilities, psychological disorders or chronic illnesses. (C) Highest educational achievement of both parents classified according to the ISCED 2011 classification 33 into three categories: primary/no education, secondary education, and university degree. As the information on children's burden was a third‐party opinion from parents, subsequent analyses were only carried out for parental burden. Results of the burden questionnaire were summarized for five age periods: toddlerhood: ≥1 to <3 years; preschool age: ≥3 to <5 years; early primary school age: ≥5 to <8 years; late primary school age: ≥8 to <12 years; secondary school age: ≥12 to <16 years. Due to small group size (n = 1 each), CPT1D and CPT2D were excluded for this analysis. The data cut‐off for this evaluation was June 7, 2022.
2.3. Statistical analysis
The perceived psychosocial burden rating was trichotomized into no burden, medium burden (including mild and medium burden), and severe burden (including severe and very severe burden). The severity of parental burden was calculated based on a score (no burden = 1, medium burden = 2, severe burden = 3). All statistical analyses were computed by R language version 4.3.2. Continuous variables are described by median and range. Statistics are based on a Bayesian approach 34 : logistic models for dichotomous outcome (e.g., no financial burden and financial burden), and cumulative ordinal regression models for ordinal outcome (e.g., no, medium or severe psychosocial burden). For each outcome, several models with different sets of predictor variables were computed. To identify valid predictors for each outcome, the models were pairwise compared by approximate leave‐one‐out cross‐validation (loo). In this procedure, the difference in expected log point‐wise predictive density (elpddiff) and its standard error (elpdse) were computed. 34 The preferred model has always elpddiff = 0 because it is the difference between the preferred model and itself, and negative values in subsequent. We used R package ‘brms’ version 2.20.4, which itself uses ‘Stan’ to compute full Bayesian statistical inference with Markov‐Chain‐Monte‐Carlo sampling (https://mc-stan.org/). R package ‘loo’ version 2.6.0 35 was used to compute approximate leave‐one‐out cross‐validation. Effects from preferred models are visualized by conditional effects plots (function “conditional_effects” in R package ‘brms’) estimating the expected means.
3. RESULTS
3.1. Study cohort
In total, 813 psychosocial visits of 369 individuals (188 female, 181 male) with 10 of the 13 IMDs of the German National Panel identified by NBS were evaluated in this study (Table 1). Until the last study visit at median age (range) of 8.1 (1–16) years, 85.9% of them showed age‐appropriate neurocognitive development (Table 1). Family financial income was low for 14.4%, medium for 32.5%, high for 41.5% and unknown for 11.7% of families. The highest level of parental education was a university degree for 29.0% of families and a secondary education for 65.0%. Overall, 39.3% of the families reported additional familial stress factors, e.g., other family members in need of care, with disabilities, psychological disorders, or chronic illnesses (Table 1).
TABLE 1.
Description of the study cohort.
| Total | BioD | GA1 | Gal | cIVA | aIVA | LCHADD/mTFPD | MCADD | MSUD | PKU | Tyr1 | cVLCADD | aVLCADD | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Number of patients (female, male) | 369 (188, 181) | 18 (4, 14) | 12 (7, 5) | 22 (12, 10) | 7 (4, 3) | 19 (8, 11) | 8 (5, 3) | 108 (51, 57) | 15 (8, 7) | 144 (80, 64) | 5 (2, 3) | 4 (3, 1) | 7 (4, 3) |
| Median (range) | |||||||||||||
| Age (at last visit) [years] | 8.1 (1.0–16.0) | 4.9 (1–15.9) | 7.9 (1.6–15.6) | 10 (1.5–14.2) | 14.1 (11.5–16.0) | 5.4 (1.3–14.4) | 5.0 (1.5–14.0) | 6.2 (1.0–15.4) | 11.3 (1.1–14.3) | 8.8 (1.0–15.9) | 13.2 (1.3–15.4) | 11.6 (4.2–15.2) | 10.3 (1.1–15.3) |
| Number | |||||||||||||
| Development/cognition (at last visit) | |||||||||||||
| Age‐appropriate | 317 | 18 | 7 | 16 | 7 | 16 | 6 | 99 | 10 | 124 | 4 | 3 | 7 |
| Non age‐appropriate | 52 | 0 | 5 | 6 | 0 | 3 | 2 | 9 | 5 | 20 | 1 | 1 | 0 |
| Dietary treatment (at last visit) | |||||||||||||
| Protein‐reduced | 164 | 0 | 11 | 0 | 6 | 3 | 0 | 0 | 13 | 128 | 3 | 0 | 0 |
| Fat‐reduced | 11 | 0 | 0 | 0 | 0 | 0 | 8 | 0 | 0 | 0 | 0 | 3 | 0 |
| Gal‐reduced | 22 | 0 | 0 | 22 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| No Diet | 171 | 18 | 1 | 0 | 1 | 16 | 0 | 108 | 2 | 16 | 1 | 1 | 7 |
| N/a | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 |
| Risk for decompensation (at last visit) | |||||||||||||
| Yes | 180 | 0 | 12 | 0 | 7 | 19 | 8 | 108 | 15 | 0 | 0 | 4 | 7 |
| No | 189 | 18 | 0 | 22 | 0 | 0 | 0 | 0 | 0 | 144 | 5 | 0 | 0 |
| Family income (at last visit) | |||||||||||||
| Low | 53 | 3 | 2 | 1 | 2 | 1 | 2 | 14 | 4 | 22 | 1 | 1 | 0 |
| Medium | 120 | 5 | 5 | 10 | 3 | 5 | 3 | 33 | 5 | 44 | 2 | 1 | 4 |
| High | 153 | 8 | 4 | 6 | 1 | 9 | 3 | 53 | 5 | 60 | 2 | 0 | 2 |
| N/a | 43 | 2 | 1 | 5 | 1 | 4 | 0 | 8 | 1 | 18 | 0 | 2 | 1 |
| Highest parental educational achievement | |||||||||||||
| Primary/no education | 5 | 0 | 0 | 0 | 0 | 0 | 0 | 2 | 1 | 1 | 0 | 1 | 0 |
| Secondary education | 240 | 6 | 6 | 16 | 6 | 14 | 7 | 71 | 10 | 91 | 5 | 2 | 6 |
| University degree | 107 | 10 | 4 | 6 | 1 | 4 | 1 | 30 | 4 | 46 | 0 | 0 | 1 |
| N/a | 17 | 2 | 2 | 0 | 0 | 1 | 0 | 5 | 0 | 6 | 0 | 1 | 0 |
| Additional familial stress factors (at last visit) | |||||||||||||
| Yes | 145 | 6 | 3 | 7 | 4 | 9 | 6 | 48 | 7 | 48 | 2 | 3 | 2 |
| No | 224 | 12 | 9 | 15 | 3 | 10 | 2 | 60 | 8 | 96 | 3 | 1 | 5 |
3.2. Psychosocial burden is associated with child's age and differs between children and parents
The probability of perceived psychosocial burden was associated with the age of the affected child and differed between children and their parents (Figure 1). In general, the overall probability of severe psychosocial burden was low for both parents and children. At younger ages, there was only mild burden for the child, whereas the probability of burden was considerably higher for the parents (Figure 1). With increasing age, the probability of burden increased for children while the parental burden tended to decrease (Figure 1).
FIGURE 1.

Psychosocial burden is associated with child's age and differs between children and parents. Probability of none (blue), medium (yellow), and severe (red) psychosocial burden for affected children and their parents [95% credibility range; model according to approximate leave‐one‐out cross‐validation].
3.3. Perceived parental burden is associated with child's diagnosis, treatment, and development
To identify associations between perceived parental burden and clinical characteristics, models including different covariates were compared (Supplementary Figures S1 and S2). The variables diagnosis, dietary treatment, and risk of decompensation were confounded, and therefore, only one of these variables could be included in the statistical model simultaneously. Model comparisons using loo criteria showed superiority of the covariate diagnosis over dietary treatment and risk of decompensation (Supplementary Figure S1). In analogy, the mean severity of parental burden averaged over age varied markedly among diagnoses (Figure 2). Diseases not requiring dietary treatment (BioD, MCADD, aIVA, aVLCADD) showed a lower severity of parental burden than those requiring strict dietary treatment (Gal, PKU, Tyr1) and posing a high risk of metabolic decompensation (MSUD, LCHADD/mTFPD) (Figure 2).
FIGURE 2.

Parental psychosocial burden differs between different diagnoses. The severity of parental burden was calculated based on a score (no burden = 1, medium burden = 2, severe burden = 3) and averaged as mean over time. It differs between diagnoses with low severity for diseases without dietary treatment (BioD, MCADD, aIVA, aVLCADD) and high severity for diseases requiring a strict dietary treatment (PKU, Tyr 1) and composing a high risk for metabolic decompensation (MSUD, LCHADD/mTFPD) [95% credibility range; model according to approximate leave‐one‐out cross‐validation].
The probability of reported burden differed between the patient groups with and without necessary dietary treatment (Figure 3). If dietary treatment was required, almost all parents experienced at least a medium burden, with a remarkably higher probability of severe burden compared to those without diet (Figure 3). No apparent differences were observed between the different diets (protein‐, galactose‐, or fat‐reduced).
FIGURE 3.

Necessity of diet is associated with higher burden. Probability of none (blue), medium (yellow), and severe (red) psychosocial burden for parents with or without necessary dietary treatment [95% credibility range; model according to approximate leave‐one‐out cross‐validation].
Subsequent analyses including further covariates (additional familial stress factors, child's neurocognitive development, and the highest parental education) revealed that the model including child development was superior to others (Supplementary Figure S2). In conclusion, parental burden was particularly associated with the diagnosis and the neurocognitive development of the child with a higher burden for parents with children with no age‐appropriate development or cognition (Figure 4).
FIGURE 4.

Parental psychosocial burden is associated with child's development and cognition. The severity of parental burden was calculated based on a score (no burden = 1, medium burden = 2, severe burden = 3) and averaged as mean over time. It differs between diagnoses and between children with age‐appropriate (blue) and delayed (red) development [95% credibility range; model according to approximate leave‐one‐out cross‐validation].
3.4. Financial burden is associated with family income and dietary treatment
At their last study visit, 24% of families reported a financial burden due to the necessary treatment of their child's IMD. A model including further covariates (age, dietary treatment, and income) revealed no significant change over time, but a clear association of financial burden with a necessary dietary treatment and, predominantly but not exclusively, low income (Figure 5).
FIGURE 5.

Financial burden is associated with family net income and diet. The probability of financial burden is associated with family net income and the necessity of a dietary treatment requirement. Net family income was classified to three categories and was adapted during the conduct of the study due to inflation, lastly: low income (<2000 €; red), medium income (2000–3500 €; yellow), and high income (> 3500€; blue) [95% credibility range; model according to approximate leave‐one‐out cross‐validation].
The negative effects of the IMD and the necessary treatment on parental professional lives showed no significant change over time and were reported by 15.5% of mothers, but only 2.9% of fathers at the last study visit, respectively (Figure 6). Analogous to parental burden, this effect was especially high for diseases with strict dietary treatment (e.g., Tyr1) and a high risk for metabolic decompensation (e.g., MSUD, LCHADD/mTFPD) (Supplementary Figure S2).
FIGURE 6.

Impact of child's diseases on maternal (red) and paternal (blue) career prospects. Mothers report a higher negative influence of their child's disease on their professional life as fathers [95% credibility range; model according to approximate leave‐one‐out cross‐validation].
3.5. Positive future outlook of parents
Regardless of their psychosocial and financial burdens, the majority of parents look ahead to a positive future expecting an independent life for their child, with a trend to a slight decrease observed once children reach school age (Supplementary Figure S3).
4. DISCUSSION
The confirmed diagnosis of a chronic disease composes a psychosocial risk. 15 , 16 Psychosocial issues related to NBS especially concerning the immediate reaction to a positive NBS result and communication with health care providers have recently been summarized. 18
However, little is known about the psychosocial and financial burdens on the long‐term development of children with an IMD identified by NBS and their parents. A previous smaller evaluation of the NBS cohort in Southwest Germany (N = 187) showed that the psychosocial burden was judged higher for parents than for their children and that financial burden was higher in families with low income and dietary treatment. 24 However, because of short follow‐up intervals, age‐dependent effects could not be investigated. 24 The present evaluation of the NBS cohort, with a substantial increase in study size (N = 369) and observation period (median 8.1 years) allowed a longitudinal assessment including age‐dependencies and clearly showed that (1) psychosocial burden reported by parents was substantial and differed in an age‐dependent way between children and their parents, (2) high disease severity, dietary treatment, and developmental delay increased the parental burden, (3) financial burden was associated with family income and diet without change over time, and (4) mainly the maternal career prospects were negatively affected by the IMD. To date, this longitudinal assessment of psychosocial and financial burden in a cohort followed for up to 16 years is unique.
4.1. Parents bear the psychosocial burden for their children at younger ages
The psychosocial burden was rated higher for the parents at younger ages of the affected child. This is a commonly reported observation. 17 , 36 , 37 , 38 In young children affected by a chronic disease, parents bear full responsibility for their child's daily treatment. Trying to make the child's life as easy as possible by overtaking as much responsibility and burden as possible might compose a parental coping mechanism for the affected child. At the same time, this might be an origin of substantial burden for the parents and underlines the need of psychosocial support for families at an early age and especially immediately after diagnosis. 18 As children grow up, they are asked to take more responsibility for their disease (if feasible) and start to perceive more clearly the disease‐related limitations and burdens, such as dietary restrictions, while parents may feel less burden due to their experience and shared responsibility. Regular psychosocial support, not only for parents, but also for adolescents and young adults with IMDs is important and should be integrated into regular metabolic consultations.
Further analyses showed that the perceived parental burden was associated with disease severity, dietary treatment, and developmental delay. It has been shown that dietary treatment increases the risk of psychosocial burden. 15 , 24 , 27 In the present study, parental burden was specifically high for diseases with strict dietary treatment (e.g., PKU, Tyr1) and high risk of decompensation (e.g., LCHADD/mTFPD, MSUD). In some of these diseases (e.g., LCHADD/mTFPD), even excellent therapy adherence did not completely prevent the onset of disease‐specific symptoms. 19 Still, the majority of patients (85.9%) included in this study showed an age‐appropriate development, confirming a previous result 13 and corresponding to the expected normal distribution of intelligence in the general population. Association of parental burden with child's development is not surprising and has been shown before, 26 but still reflects that these family are in need for additional support.
4.2. Financial burden is a great source of stress for families
The distribution of the family net income slightly differed from that in the general population. Specifically, a smaller percentage of families reported low income compared to the general population in Germany (general population: low 30.7%, medium 32.2%, high 37.1% 39 ; study group: low 14.4%, medium 32.5%, high 41.5%, n/a 11.7%). However, this comparison is limited by the percentage of families in our study who did not disclose their family income.
The financial burden of families was not age‐dependent and especially common in families with low income and a necessary dietary treatment which should normally not be expected in a highly developed European country with a solid health insurance program. 40 Amino acid mixtures are an essential component of balanced metabolic diets and are usually covered by health insurance in Germany. 41 However, other necessary dietary products that cost up to fourfold more than regular foods, such as low‐protein or lactose‐free food are rarely paid by the insurance companies in Germany regardless of age. 42 , 43 Financial stress has been stated before as a great sources of stress for families with a child with MSUD, 25 UCD, 22 or other IMDs. 44 , 45 , 46 Exemplarily, the financial extra cost for a Phe‐restricted diet has been calculated to be 604€ per year in 2013. 47 As financial limitations might have a negative impact on treatment adherence and the clinical outcome, more attention for families and counseling regarding financial support is necessary.
4.3. Negative impact on maternal career prospects
A child with an IMD mostly affected the maternal career prospects, confirming a major involvement of women as primary caregivers, which is in line with previous studies. 17 , 46 , 48 , 49 No age‐dependent effect was observed, affirming that this gender gap cannot be attributed solely to parental leave taken by mothers directly after childbirth.
Regarding future expectations for an independent life, there is a slight decrease in parents' expectations just after the beginning of school age, anticipating their child's need for special support, possibly indicating the time when a developmental delay becomes obvious.
4.4. Further development of NBS in the light of perceived burdens
Overall, NBS is a major achievement in secondary disease prevention in individuals with an IMD and is the prerequisite for a favorable long‐term health outcome. 9 , 10 , 11 , 13 , 14 Despite the described psychosocial burdens, it has been shown before that parents of children diagnosed through NBS experience lower levels of stress compared to parents of clinically diagnosed children. 26 , 50 , 51 Pilot projects to test for candidate metabolic disorders 52 , 53 , 54 , 55 as well as the increasing availability of whole genome sequencing as the next technical milestone in the further development of NBS 56 , 57 , 58 will lead to further expansion of NBS programs. At first glance, common strategies such as dietary treatment might appear as a low‐risk and low‐cost therapy compared to some innovative and more advanced therapies. However, from the perspective of parents having a child with an IMD, this daily and often life‐long intervention turns out to be complex and burdensome and to put substantial psychosocial and financial burdens on them. 15 , 22 , 24 , 25 , 27 , 47 Clear‐cut case definitions and indications to treat are required to minimize this burden and to prevent unnecessary intervention. 59 For example, individuals with aIVA identified by NBS so far have not developed metabolic decompensations or disease‐specific symptoms, and the metabolic therapy (low‐protein diet, medication) previously conducted for many years in analogy to that of individuals with the severe cIVA variant turned out to be probably dispensable. 7 , 60 , 61 Therefore, the core requirements of all present and future NBS programs include a clear and regularly updated case definition as well as evidence‐based treatment indications and recommendations, such as for GA1 patients. 62
4.5. Limitations
The questionnaire is designed to allow efficient completion during routine outpatient visits without the need for an extra appointment with trained staff, and thereby accepting the limitations of a non‐standardized format. Data are based on parents' subjective perception only with children's burden being always assessed indirectly by questioning their parents. Therefore, parent's perspective of child's burden could not be compared to the child's own perspective, even not for adolescents. Due to this limitation, further analyses have focused on perceived parental burden. Additionally, the study design did not include a control group (e.g., healthy or symptomatically diagnosed children). Further research, including longitudinal assessments using standardized tests or questionnaires, is needed to capture and further investigate all sources of burden in detail, and subsequently to improve the socio‐economic situation of families with a child diagnosed with an IMD by NBS.
5. CONCLUSION
A remarkable psychosocial and financial burden has been shown for parents having a child with an IMD. The perceived burdens were associated with the child's age, disease, mode of treatment and development, and family income. Therefore, regular psychosocial and financial counseling should be a mandatory part of metabolic outpatient visits. Moreover, psychological, social, and financial services should be strengthened within health programs.
AUTHOR CONTRIBUTIONS
Elena Schnabel‐Besson, Ulrike Mütze, and Stefan Kölker conceptualized and designed the study. Elena Schnabel‐Besson collected clinical data including completion of the burden questionnaire, carried out statistical analyses, evaluated and interpreted data, and drafted the initial manuscript. Ulrike Mütze and Stefan Kölker coordinated and supervised data collection, evaluated and interpreted data, and drafted the manuscript. Sven F. Garbade carried out statistical analyses and critically reviewed and revised the manuscript. Florian Gleich carried out data collection and management, and critically reviewed and revised the manuscript. Sarah C. Grünert, Julia B. Hennermann, Peter Freisinger, Johannes Krämer, and Eva Thimm collected the comprehensive follow‐up data at their study sites and critically reviewed and revised the manuscript. Gwendolyn Gramer, Georg F. Hoffmann, Marina A. Morath, and A. Tunç Tuncel collected clinical data and critically reviewed and revised the manuscript. Peter Burgard and Svenja Keßler carried out psychological testing, collected clinical data including completion of the burden questionnaire, and critically reviewed and revised the manuscript. The corresponding author (Ulrike Mütze) had full access to the complete dataset of the study and had final responsibility for the decision to submit for publication. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
FUNDING INFORMATION
The pilot study Newborn screening 2020/2025 (NGS2025) and the long‐term outcome study (Long‐term outcome of patients with inherited metabolic diseases after diagnosis by expanded newborn screening) are generously supported by the Dietmar Hopp Foundation, St. Leon‐Rot, Germany (grant numbers 23011220, 23011221, 1DH1911376, and 1DH2011117 to SK and GFH).
CONFLICT OF INTEREST STATEMENT
Stefan Kölker and Georg F. Hoffmann reported grants from the Dietmar Hopp Foundation (St Leon‐Rot, Germany) during the conduct of the study. Ulrike Mütze reported a grant from the Medical Faculty of Heidelberg University, during the conduct of the study. Sarah C. Grünert reported fees from Ultragenyx GmbH, Vitaflo GmbH, and Danone GmbH, outside the submitted work. Johannes Krämer reported personal fees from Biomarin and Sanofi, outside the submitted work. Julia B. Hennermann reported grants and fees from Sanofi, Takeda, Amicus, and Chiesi, grants from Biomarin, and fees from Immedica, outside the submitted work. Peter Freisinger, Gwendolyn Gramer, Elena Schnabel‐Besson, Sven F. Garbade, Eva Thimm, Peter Burgard, Marina A. Morath, A. Tunç Tuncel, Svenja Keßler, and Florian Gleich have nothing to disclose.
The authors confirm their independence from the sponsors; the study design and the content of the article have not been influenced by the sponsors.
ETHICS STATEMENT
The long‐term outcome study was approved by the Ethics Committee of the Medical Faculty of Heidelberg University (Vote S‐104/2005) and consecutively by all study sites.
PATIENT CONSENT STATEMENT
Informed consent was obtained from all subjects involved in the study. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000 (5).
Supporting information
Supplementary Figure S1. Socio‐economic questionnaire. This questionnaire is completed at each study visit by parents of children with an IMD identified by NBS and enrolled in the longitudinal observational NBS outcome study. The selection of questions relevant to this evaluation is shown.
Supplementary Figure S2. Negative impact of the child's diagnosis on maternal career prospects [95% credibility range; model according to approximate leave‐one‐out cross‐validation].
Supplementary Figure S3. Parental expectancy of an independent life for their child [95% credibility range; model according to approximate leave‐one‐out cross‐validation].
Supplementary Table S1. Diagnosis model is superior. Comparison of models created using loo (approximate leave‐one‐out cross‐validation) without further covariates. The model with diagnosis is superior to the other models [elpddiff: difference in expected log point‐wise predictive density, elpdse: standard error of expected log point‐wise predictive density].
Supplementary Table S2. Model including diagnosis and child's development is superior. Comparison of models including diagnosis and other psychosocial covariates created using loo (approximate leave‐one‐out cross‐validation). The model with diagnosis and development is superior to the other models [elpddiff: difference in expected log point‐wise predictive density, elpdse: standard error of expected log point‐wise predictive density].
ACKNOWLEDGMENTS
The authors are deeply indebted to the patients and their families for their participation and trust. Many thanks to all the colleagues who participated throughout the years through continuous interaction and exchange with the screening center as well as by providing information on their patients, especially named should be Junmin Fang‐Hoffmann, Dorothea Haas, Friederike Hörster, and Thomas Opladen. This extensive study over two decades was only made possible by the continuous and generous support of the Dietmar Hopp Foundation, St. Leon‐Rot, Germany. Open Access funding enabled and organized by Projekt DEAL.
Schnabel‐Besson E, Garbade SF, Gleich F, et al. Parental and child's psychosocial and financial burden living with an inherited metabolic disease identified by newborn screening. J Inherit Metab Dis. 2025;48(1):e12784. doi: 10.1002/jimd.12784
Stefan Kölker and Ulrike Mütze contributed equally to this study.
REFERENCES
- 1. Schulze A, Lindner M, Kohlmüller D, Olgemöller K, Mayatepek E, Hoffmann GF. Expanded newborn screening for inborn errors of metabolism by electrospray ionization‐tandem mass spectrometry: results, outcome, and implications. Pediatrics. 2003;111:1399‐1406. [DOI] [PubMed] [Google Scholar]
- 2. Wilcken B, Wiley V, Hammond J, Carpenter K. Screening newborns for inborn errors of metabolism by tandem mass spectrometry. N Engl J Med. 2003;348:2304‐2312. [DOI] [PubMed] [Google Scholar]
- 3. Richtlinie zur Früherkennung von Krankheiten bei Kindern, G.B.d.Ä.u. Krankenkassen, Editor. 2022, Gemeinsamer Bundesausschuss der Ärzte und Krankenkassen: https://www.g-ba.de/downloads/62-492-2848/Kinder-RL_2022-04-21_iK-2022-06-23.pdf. Accessed February 01 2024.
- 4. Boy N, Mengler K, Heringer‐Seifert J, Hoffmann GF, Garbade SF, Kölker S. Impact of newborn screening and quality of therapy on the neurological outcome in glutaric aciduria type 1: a meta‐analysis. Genet Med. 2021;23:13‐21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Boy N, Mengler K, Thimm E, et al. Newborn screening: a disease‐changing intervention for glutaric aciduria type 1. Ann Neurol. 2018;83:970‐979. [DOI] [PubMed] [Google Scholar]
- 6. Märtner EMC, Thimm E, Guder P, et al. The biochemical subtype is a predictor for cognitive function in glutaric aciduria type 1: a national prospective follow‐up study. Sci Rep. 2021;11:19300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Mütze U, Henze L, Gleich F, et al. Newborn screening and disease variants predict neurological outcome in isovaleric aciduria. J Inherit Metab Dis. 2021;44:857‐870. [DOI] [PubMed] [Google Scholar]
- 8. Mengler K, Garbade SF, Gleich F, et al. Treatment outcomes for maple syrup urine disease detected by newborn screening. Pediatrics. 2024;154(2):e2023064370. [DOI] [PubMed] [Google Scholar]
- 9. Couce ML, Castiñeiras DE, Bóveda MD, et al. Evaluation and long‐term follow‐up of infants with inborn errors of metabolism identified in an expanded screening programme. Mol Genet Metab. 2011;104:470‐475. [DOI] [PubMed] [Google Scholar]
- 10. Landau YE, Waisbren SE, Chan LM, Levy HL. Long‐term outcome of expanded newborn screening at Boston children's hospital: benefits and challenges in defining true disease. J Inherit Metab Dis. 2017;40:209‐218. [DOI] [PubMed] [Google Scholar]
- 11. Lindner M, Gramer G, Haege G, et al. Efficacy and outcome of expanded newborn screening for metabolic diseases‐‐report of 10 years from South‐West Germany. Orphanet J Rare Dis. 2011;6:44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Mütze U, Garbade SF, Gleich F, et al. Long‐term anthropometric development of individuals with inherited metabolic diseases identified by newborn screening. J Inherit Metab Dis. 2023;46:15‐27. [DOI] [PubMed] [Google Scholar]
- 13. Mütze U, Garbade SF, Gramer G, et al. Long‐term outcomes of individuals with metabolic diseases identified through newborn screening. Pediatrics. 2020;146(5):e20200444. [DOI] [PubMed] [Google Scholar]
- 14. Wilcken B, Haas M, Joy P, et al. Expanded newborn screening: outcome in screened and unscreened patients at age 6 years. Pediatrics. 2009;124:e241‐e248. [DOI] [PubMed] [Google Scholar]
- 15. Bösch F, Landolt MA, Baumgartner MR, et al. Health‐related quality of life in paediatric patients with intoxication‐type inborn errors of metabolism: analysis of an international data set. J Inherit Metab Dis. 2021;44:215‐225. [DOI] [PubMed] [Google Scholar]
- 16. Hysing M, Elgen I, Gillberg C, Lie SA, Lundervold AJ. Chronic physical illness and mental health in children. Results from a large‐scale population study. J Child Psychol Psychiatry. 2007;48:785‐792. [DOI] [PubMed] [Google Scholar]
- 17. Thiel M, Garbade SF, Rosenbaum‐Fabian S, Spiekerkoetter U, Grünert SC. Psychosocial issues and coping strategies in families affected by long‐chain fatty acid oxidation disorders. JIMD Reports. 2024;65:25‐38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Tluczek A, Ersig AL, Lee S. Psychosocial issues related to newborn screening: a systematic review and synthesis. Int J Neonatal Screen. 2022;8(4):53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Mütze U, Ottenberger A, Gleich F, et al. Neurological outcome in long‐chain hydroxy fatty acid oxidation disorders. Ann Clin Transl Neurol. 2024;11(4):883–898. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Reischl‐Hajiabadi AT, Schnabel E, Gleich F, et al. Outcomes after newborn screening for propionic and methylmalonic acidemia and homocystinurias. J Inherit Metab Dis. 2024;47:674‐689. [DOI] [PubMed] [Google Scholar]
- 21. Bush L, Davidson H, Gelles S, Lea D, Koehly LM. Experiences of families caring for children with newborn screening‐related conditions: implications for the expansion of genomics in population‐based neonatal public health programs. Int J Neonatal Screen. 2022;8(2):35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Cederbaum JA, LeMons C, Rosen M, Ahrens M, Vonachen S, Cederbaum SD. Psychosocial issues and coping strategies in families affected by urea cycle disorders. J Pediatr. 2001;138:S72‐S80. [DOI] [PubMed] [Google Scholar]
- 23. Curie A, Lion‐François L, Valayannopoulos V, et al. Clinical characteristics, developmental trajectory, and caregiver burden of patients with creatine transporter deficiency (SLC6A8). Neurology. 2024;102:e209243. [DOI] [PubMed] [Google Scholar]
- 24. Gramer G, Haege G, Glahn EM, Hoffmann GF, Lindner M, Burgard P. Living with an inborn error of metabolism detected by newborn screening‐parents' perspectives on child development and impact on family life. J Inherit Metab Dis. 2014;37:189‐195. [DOI] [PubMed] [Google Scholar]
- 25. Packman W, Henderson SL, Mehta I, et al. Psychosocial issues in families affected by maple syrup urine disease. J Genet Couns. 2007;16:799‐809. [DOI] [PubMed] [Google Scholar]
- 26. Waisbren SE, Rones M, Read CY, Marsden D, Levy HL. Brief report: predictors of parenting stress among parents of children with biochemical genetic disorders. J Pediatr Psychol. 2004;29:565‐570. [DOI] [PubMed] [Google Scholar]
- 27. Zeltner NA, Landolt MA, Baumgartner MR, et al. Living with intoxication‐type inborn errors of metabolism: a qualitative analysis of interviews with paediatric patients and their parents. JIMD Reports. 2017;31:1‐9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. van den Heuvel LM, van der Pal SM, Verschoof‐Puite RK, et al. Psychosocial impact of a true‐positive, false‐positive, or inconclusive newborn bloodspot screening result: a questionnaire study among parents. Int J Neonatal Screen. 2024;10(1):18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. AWMF . Konfirmationsdiagnostikbei Verdacht auf angeborene Stoffwechselkrankheiten aus demNeugeborenenscreening, 027‐021. https://register.awmf.org/de/leitlinien/detail/027-021. Accessed February 01, 2024. [Google Scholar]
- 30. G‐BA . Kinder‐Richtlinie: Screening von Neugeborenen zur Früherkennung der Tyrosinämie Typ I mittels Tandem‐Massenspektrometrie. Book Kinder‐Richtlinie: Screening von Neugeborenen zur Früherkennung der Tyrosinämie Typ I mittels Tandem‐Massenspektrometrie; 2018. [Google Scholar]
- 31. Liebig M, Schymik I, Mueller M, et al. Neonatal screening for very long‐chain acyl‐coA dehydrogenase deficiency: enzymatic and molecular evaluation of neonates with elevated C14:1‐carnitine levels. Pediatrics. 2006;118:1065‐1069. [DOI] [PubMed] [Google Scholar]
- 32. Statista. de.statista.com. 2024. https://de.statista.com/Statistisches_Bundesamt. Statista. Accessed February01, 2024.
- 33.ISCED, U. 2011. International standard classification of education. United Nations Educational, Scientific and Cultural Organization (UNESCO) .
- 34. Vehtari A, Gelman A, Gabry J. Practical Bayesian model evaluation using leave‐one‐out cross‐validation and WAIC. Statistics and Computing. 2017;27:1413‐1432. [Google Scholar]
- 35. Vehtari A, Gabry J, Magnusson M, et al. 2023. loo: Efficient leave‐one‐out cross‐validation and WAIC for Bayesian models. R package version 2.4. 1. https://mc-stan.org/loo/.
- 36. Fidika A, Salewski C, Goldbeck L. Quality of life among parents of children with phenylketonuria (PKU). Health Qual Life Outcomes. 2013;11:54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Scharping M, Brennenstuhl H, Garbade SF, et al. Unmet Needs of Parents of Children With Urea Cycle Disorders. Children; 2022. 9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Siddiq S, Wilson BJ, Graham ID, et al. Experiences of caregivers of children with inherited metabolic diseases: a qualitative study. Orphanet J Rare Dis. 2016;11:168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Statista 2024. Verteilung der Privathaushalte in Deutschland nach monatlichemHaushaltsnettoeinkommen im Jahr 2023. Statistisches_Bundesamt. https://de.statista.com/statistik/daten/studie/3048/umfrage/privathaushalte‐nach‐monatlichem‐haushaltsnettoeinkommen/. Accessed June 01, 2024.
- 40. Blümel M, Spranger A, Achstetter K, Maresso A, Busse R. Germany: health system review. Health Syst Transit. 2020;22:1‐272. [PubMed] [Google Scholar]
- 41.G‐BA (2018). Beschluss Kinder‐Richtlinie: Screening von Neugeborenen zur Früherkennung von SCID. G. B. d. Ä. u. Krankenkassen. BAnz AT 08.02.2019 B2. https://www.g-ba.de/beschluesse/3586/, Gemeinsamer Bundesausschuss der Ärzte und Krankenkassen.
- 42. Berry SA, Kenney MK, Harris KB, et al. Insurance coverage of medical foods for treatment of inherited metabolic disorders. Genet Med. 2013;15:978‐982. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. LSG_Baden‐Württemberg. 2010. L 5 KR 2103/09. S. B. Deutschland. https://www.sozialgerichtsbarkeit.de/legacy/132259. Accessed June 01, 2024.
- 44. Gündüz M, Yüksel Güdek Y, Kasapkara ÇS. Out‐of‐pocket health expenditures in patients living with ınborn errors of metabolism. Orphanet J Rare Dis. 2023;18:179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Mlčoch T, Puda R, Ješina P, Lhotáková M, Štěrbová Š, Doležal T. Dietary patterns, cost and compliance with low‐protein diet of phenylketonuria and other inherited metabolic diseases. Eur J Clin Nutr. 2018;72:87‐92. [DOI] [PubMed] [Google Scholar]
- 46. Sestini S, Paneghetti L, Lampe C, et al. Social and medical needs of rare metabolic patients: results from a MetabERN survey. Orphanet J Rare Dis. 2021;16:336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Eijgelshoven I, Demirdas S, Smith TA, van Loon JM, Latour S, Bosch AM. The time consuming nature of phenylketonuria: a cross‐sectional study investigating time burden and costs of phenylketonuria in The Netherlands. Mol Genet Metab. 2013;109:237‐242. [DOI] [PubMed] [Google Scholar]
- 48. Chu SY, Wen CC, Weng CY. Gender differences in caring for children with genetic or rare diseases: A mixed‐methods study. Children; 2022:9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Walkowiak D, Domaradzki J, Mozrzymas R, et al. Professional activity, gender and disease‐related emotions: the impact on parents' experiences in caring for children with phenylketonuria. Mol Genet Metab Rep. 2023;36:100992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Boychuk NA, Mulrooney NS, Kelly NR, Goldenberg AJ, Silver EJ, Wasserstein MP. Parental depression and anxiety associated with newborn bloodspot screening for rare and variable‐onset disorders. Int J Neonatal Screen. 2022;8(4):59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Waisbren SE, Albers S, Amato S, et al. Effect of expanded newborn screening for biochemical genetic disorders on child outcomes and parental stress. JAMA. 2003;290:2564‐2572. [DOI] [PubMed] [Google Scholar]
- 52. Maier EM, Mütze U, Janzen N, et al. Collaborative evaluation study on 18 candidate diseases for newborn screening in 1.77 million samples. J Inherit Metab Dis. 2023;46:1043‐1062. [DOI] [PubMed] [Google Scholar]
- 53. Martín‐Rivada Á, Cambra Conejero A, Martín‐Hernández E, et al. Newborn screening for propionic, methylmalonic acidemia and vitamin B12 deficiency. Analysis of 588,793 newborns. J Pediatr Endocrinol Metab. 2022;35:1223‐1231. [DOI] [PubMed] [Google Scholar]
- 54. Pajares S, Arranz JA, Ormazabal A, et al. Implementation of second‐tier tests in newborn screening for the detection of vitamin B(12) related acquired and genetic disorders: results on 258,637 newborns. Orphanet J Rare Dis. 2021;16:195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55. Schnabel E, Kölker S, Gleich F, et al. Combined newborn screening allows comprehensive identification also of attenuated phenotypes for methylmalonic acidurias and homocystinuria. Nutrients. 2023;15(15):3355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56. Kingsmore SF, Smith LD, Kunard CM, et al. A genome sequencing system for universal newborn screening, diagnosis, and precision medicine for severe genetic diseases. Am J Hum Genet. 2022;109:1605‐1619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57. Schaaf CP, Kölker S, Hoffmann GF. Genomic newborn screening: proposal of a two‐stage approach. J Inherit Metab Dis. 2021;44:518‐520. [DOI] [PubMed] [Google Scholar]
- 58. Stark Z, Scott RH. Genomic newborn screening for rare diseases. Nat Rev Genet. 2023;24:755‐766. [DOI] [PubMed] [Google Scholar]
- 59. Mütze U, Mengler K, Boy N, et al. How longitudinal observational studies can guide screening strategy for rare diseases. J Inherit Metab Dis. 2022;45:889‐901. [DOI] [PubMed] [Google Scholar]
- 60. Ensenauer R, Vockley J, Willard JM, et al. A common mutation is associated with a mild, potentially asymptomatic phenotype in patients with isovaleric acidemia diagnosed by newborn screening. Am J Hum Genet. 2004;75:1136‐1142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61. Vockley J, Ensenauer R. Isovaleric acidemia: new aspects of genetic and phenotypic heterogeneity. Am J Med Genet C Semin Med Genet. 2006;142c:95‐103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62. Boy N, Mühlhausen C, Maier EM, et al. Recommendations for diagnosing and managing individuals with glutaric aciduria type 1: third revision. J Inherit Metab Dis. 2023;46:482‐519. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Figure S1. Socio‐economic questionnaire. This questionnaire is completed at each study visit by parents of children with an IMD identified by NBS and enrolled in the longitudinal observational NBS outcome study. The selection of questions relevant to this evaluation is shown.
Supplementary Figure S2. Negative impact of the child's diagnosis on maternal career prospects [95% credibility range; model according to approximate leave‐one‐out cross‐validation].
Supplementary Figure S3. Parental expectancy of an independent life for their child [95% credibility range; model according to approximate leave‐one‐out cross‐validation].
Supplementary Table S1. Diagnosis model is superior. Comparison of models created using loo (approximate leave‐one‐out cross‐validation) without further covariates. The model with diagnosis is superior to the other models [elpddiff: difference in expected log point‐wise predictive density, elpdse: standard error of expected log point‐wise predictive density].
Supplementary Table S2. Model including diagnosis and child's development is superior. Comparison of models including diagnosis and other psychosocial covariates created using loo (approximate leave‐one‐out cross‐validation). The model with diagnosis and development is superior to the other models [elpddiff: difference in expected log point‐wise predictive density, elpdse: standard error of expected log point‐wise predictive density].
