Abstract
Autism spectrum disorder (ASD) is a neurodevelopmental disorder that results from a complex interaction between genes and environment. Breastfeeding (BF) is thought to promote healthy cognitive development, and a body of research has suggested that it may also protect against ASD. Our objectives were to identify the relationship between the initiation and duration of BF and the severity of clinical presentation in ASD. Data were collected from 243 children with a confirmed diagnosis of ASD followed in the ELENA cohort. Clinical severity was measured according to multiple dimensions using standardised tools. The frequency of the initiation of BF was comparable to that of the general population and the rate of children still being breastfed at six months of age was higher. Our results did not indicate a contribution of initiation or duration of BF to the prevention of clinical severity of ASD. We discuss our results in the light of possible methodological limitations of previous reports of an association between BF and ASD.
Clinical Trial Registration: NCT02625116.
Introduction
Autism spectrum disorder (ASD) is a persistent neurodevelopmental disability. It is clinically evident from early childhood and related to brain development1. The characteristics of clinical presentation can be very heterogeneous in terms of both severity and outcome2. The risk is believed to be multifactorial, related to an interplay of genetic and environmental factors during critical perinatal windows. A growing number of studies are focusing on the nutritional status of children with ASD3–5.
Maternal malnutrition during pregnancy has been found to be associated with an increased risk of adverse neurodevelopmental outcomes in offspring6–8. There is also evidence of an association between breastfeeding (BF) and better cognitive outcome9–11, particularly when BF is exclusive10–14. This positive impact of BF could be due to the nutritional or hormonal content of breast-milk or to the social contact between mother and child during the act of nursing15. However, conflicting literature exists regarding the consistency of this positive impact attributed to BF after adjustment for important covariates such as parental education and socio-economic status. This suggests caution in the interpretation of these results16.
Studies on the association between BF and ASD show conflicting results. Two recent meta-analyses suggest that breastfeeding is protective against ASD17,18. Ghozy et al.17 found a reduction in the risk of ASD of 58% with ever breastfeeding, 76% with ever exclusive breastfeeding and 54% with extended breastfeeding for 6 months.
Tseng et al.18 also reviewed 7 articles reporting an association between breastfeeding and an ASD diagnosis. They found that children with ASD were significantly less likely to have been breastfed than those without (OR = 0.61, 95% CI = [0.45–0.83], P = 0.002).
However, another part of the literature qualifies the finding of an association between BF and ASD. Husk and Keim19 examined data from the National Survey of Children's Health from 2007 to 2011—a nationally representative survey of US children—to determine whether infant breastfeeding was associated with the later development of ASD. Approximately 38,000 respondents were asked (i) whether their child currently had ASD and (ii) to recount their child's breastfeeding history. A series of covariate-controlled analyses of the NSCH indicated that a current diagnosis of ASD (n = 391) was unassociated with any measure of BF history. The authors speculate that previous reports of such associations might reflect methodological limitations. Another study of 673 children diagnosed with ASD and 876 controls20 found that the initiation of BF was not associated with ASD after adjustment for child and mother socioeconomic and pregnancy characteristics. However, in this study, the mothers of children with ASD were less likely to report long duration of BF. This contrasts with results of other studies21,22 which found that longer duration of breastfeeding protects against autistic traits or ASD diagnoses.
There is an important heterogeneity in the clinical presentation of ASD related to intellectual abilities, symptom severity and behavioural patterns, these factors all being involved in the outcome trajectories2,23. To date, no study has investigated whether breastfeeding is also a protective factor for the severity of the clinical presentation of children diagnosed with ASD.
Our aim was to investigate the association between first the initiation of breastfeeding and clinical presentation severity in ASD, and second the duration of any and predominant breastfeeding and clinical presentation severity in ASD.
Results
Study sample
Among the 243 children included, aged 10.3 years (± 3.9), 81.9% (n = 199) were males. At inclusion of participants in the ELENA cohort, their mean ADOS-2 CSS score was 6.8 (± 2.0), their mean IQ was 77.1 (± 28.0) and their SRS-2 T-score was 94.9 (± 19.0). Their mean VABS-II scores were 72.6 (± 15.5) for communication, 75.6 (± 12.8) for daily living skills and 70.9 (± 10.6) for socialisation. Their mean ABC scores were 33.1 (± 21.0) for irritability/ aggressiveness, 26.6 (± 18.4) for lethargy/ social withdrawal, 29.6 (± 22.9) for stereotyped behaviours/ self-harm and 44.0 (± 25.0) for hyperactivity/ lack of cooperation. On average, children included in the analyses had higher IQ, VABS-II daily living skills and communication scores than in the whole ELENA cohort (see Supplementary Table S2 online).
Birth conditions, family and socio-demographic characteristics
Of the 243 children in our sample, at the time of birth, 93% had parents (n = 226) who were living together and 53% did not have any siblings. The average age of the mothers was 31.8 years. For about 36% (n = 87), at least one of the parents had a university degree or more. Of the respondents, 72.8% (n = 177) had been breastfed. Breastfed children had parents with higher levels of education and SES (P < 0.001 and P = 0.02, respectively) than non-breastfed children, and they were more often only children (P < 0.001) (Table 1).
Table 1.
Characteristics of participants at the time of birth.
| All | Breastfed | Never-Breastfed | P value | |
|---|---|---|---|---|
| (n = 243) | (n = 177) | (n = 66) | ||
| Children’s characteristics | ||||
| Sex | ||||
| Boy | 199 (81.9%) | 142 (80.2%) | 57 (86.4%) | 0.27 |
| Girl | 44 (18.1%) | 35 (19.8%) | 9 (13.6%) | |
| Gestational age | ||||
| Preterm < 37 weeks | 21 (9.3%)h | 13 (8.0%)g | 8 (12.5%)b | 0.29 |
| Term ≥ 37 weeks | 206 (90.7%) | 150 (92.0%) | 56 (87.5%) | |
| Delivery mode | ||||
| Vaginal | 194 (79.8%) | 137 (77.4%) | 57 (86.4%) | 0.12 |
| Caesarean section | 49 (20.2%) | 40 (22.6%) | 9 (13.6%) | |
| Mother’s characteristics | ||||
| Age at delivery (years) | 31.77 (± 5.35) | 31.44 (± 5.12) | 32.67 (± 5.88) | 0.13 |
| Obese before pregnancy (yes)i | 22 (9.4%)f | 13 (7.6%)e | 9 (14.3%)c | 0.12 |
| Body Mass Index before pregnancy (kg/m2) | 22.6 (20.0–25.6)f | 22.6 (20.0–25.6)e | 22.2 (20.0–25.6)c | 0.89 |
| Tobacco smoking during pregnancy (yes) | 38 (15.8%)c | 26 (14.9%)b | 12 (18.5%)a | 0.5 |
| Socio-environmental characteristics | ||||
| Parental education levelj | ||||
| Below high-school diploma | 19 (7.9%)a | 8 (4.5%)a | 11 (16.7%) | < 0.001 |
| High school to 2-year university degree | 136 (56.2%) | 89 (50.6%) | 47 (71.2%) | |
| 3-year university degree | 87 (35.9%) | 79 (44.9%) | 8 (12.1%) | |
| Parental socioeconomic status (SES)k | ||||
| Low SES | 81 (33.6%)b | 50 (28.4%)a | 31 (47.7%)a | 0.02 |
| Middle SES | 81 (33.6%) | 65 (36.9%) | 16 (24.6%) | |
| High SES | 79 (32.8%) | 61 (34.7%) | 18 (27.7%) | |
| Number of older siblings | ||||
| 0 | 127 (53.1%)d | 101 (58.0%)c | 26 (40.0%)a | < 0.001 |
| 1 | 68 (28.5%) | 53 (30.5%) | 15 (23.1%) | |
| ≥ 2 | 44 (18.4%) | 20 (11.5%) | 24 (36.9%) | |
| Marital status at birth | ||||
| Single mother | 17 (7.0%) | 13 (7.3%) | 4 (6.1%) | 0.73 |
| Parents living together | 226 (93.0%) | 164 (95.7%) | 62 (93.9%) | |
| Professional status 6 months after birth | ||||
| Unemployed | 142 (58.4%) | 98 (55.4%) | 44 (66.7%) | 0.11 |
| Employed or involved in training/studies | 101 (41.6%) | 79 (44.6%) | 22 (33.3%) | |
Data are presented as the mean (± SD), median (IQR) or n (%).
a1 missing value.
b2 missing values.
c3 missing values.
d4 missing values.
e5 missing values.
f8 missing values.
g14 missing values.
h16 missing values.
iObese: body mass index ≥ 30 kg/m2.
jParental education: highest level of education of one of the parents.
kLow SES: farm workers, labourers, service workers and unemployed; Medium SES: farmers, supervisors, skilled craftsmen; High SES: business owners, professionals, managers.
Breastfeeding conditions
The median duration of ABF was 6 months (IQR = [2.7–9.0]), with 47.5% of children breastfed for less than 6 months (n = 84), 34.5% (n = 61) for 6–12 months and 18% (n = 32) for more than 12 months. The median duration of PBF was 3.5 months (IQR = [1.0–5.0]), with 78% of children mainly breastfed for less than 6 months (n = 138).
Association between breastfeeding and severity of clinical presentation
Breastfed children had higher VABS-II scores for daily living skills than those who had never been breastfed (77.1 (± 12.7) vs. 71.5 (± 12.2), respectively, p = 0.002). They also tended to have higher VABS-II communication scores (73.7 (± 15.2) vs. 69.8 (± 15.9), respectively, P = 0.08). Breastfeeding was not associated with the VABS-II socialisation score, the SRS-2 T-score, ADOS-2 CSS, IQ or the ABC scores (Table 2).
Table 2.
Association between breastfeeding initiation status and ASD clinical severity at inclusion.
| Breastfed | Never-Breastfed | P value | |
|---|---|---|---|
| Intellectual quotient | (N = 177) | (N = 66) | |
| IQ | 78.2 (± 28.0) | 73.9 (± 28.0) | 0.35 |
| VABS-II standard scores | (N = 177) | (N = 66) | |
| Communication | 73.7 (± 15.2) | 69.8 (± 15.9) | 0.08 |
| Daily living skills | 77.1 (± 12.7) | 71.5 (± 12.2) | 0.002 |
| Socialisation | 71.4 (± 11.2) | 69.7 (± 8.8) | 0.50 |
| ADOS-2 CSS | (N = 157) | (N = 55) | |
| 6.8 (± 1.9) | 6.7 (± 2.1) | 0.60 | |
| ABC | (N = 136) | (N = 49) | |
| Irritability/ aggressiveness | 31.1 (15.6–46.7)a | 31.1 (17.8–46.7) | 0.85 |
| Lethargy/ social withdrawal | 24.0 (12.5–37.5) | 27.1 (16.7–33.3) | 0.98 |
| Stereotyped behaviours/ self-harm | 23.8 (9.5–47.6) | 23.8 (9.5–61.9) | 0.66 |
| Hyperactivity/ lack of cooperation | 43.8 (25.0–65.6) | 41.7 (14.6–58.3) | 0.21 |
| SRS-2 T-score | (N = 110) | (N = 38) | |
| 94.9 (± 18.8) | 95.0 (± 19.7) | 0.97 |
Data are presented as the mean (± SD) or median (IQR).
VABS-II vineland second version, ADOS-2 CSS autism diagnostic observation schedule second version calibrate severity scale, ABC Aberrant Behavior Checklist, SRS-2 Social-Responsiveness Scale, second version.
a1 missing value.
Significant values are in bold.
After adjustment for potential confounders, breastfeeding was not significantly associated with VABS-II scores for daily living skills (P = 0.09) or communication (P = 0.76) (Table 3).
Table 3.
Adjusted multiple linear regression models to explain VABS-II Daily living skills and communication scores (n = 243).
| VABS-II Daily living skills | VABS-II Communication | |||||
|---|---|---|---|---|---|---|
| R2 | β (95% IC) | P value | R2 | β (95% IC) | P value | |
| 0.12 | 0.11 | |||||
| Mothers | ||||||
| Age (years) | 0.15 (− 0.17; 0.46) | 0.36 | 0.31 (− 0.07; 0.69) | 0.11 | ||
| Tobacco smoking during pregnancyd (yes) | 1.24 (− 3.21; 5.69) | 0.59 | 0.87 (− 4.53; 6.26) | 0.75 | ||
| Obesity before pregnancya,d (yes) | 0.79 (− 5.41; 6.99) | 0.8 | − 5.28 (− 12.38; 1.82) | 0.15 | ||
| Child | ||||||
| Girl | − 3.82 (− 8.03; 0.38) | 0.07 | − 0.77 (− 5.87; 4.33) | 0.77 | ||
| Gestational aged (ref : ≥ 37 weeks) | ||||||
| < 37 weeks | − 0.16 (− 5.72 ; 5.40) | 0.96 | − 3.41 (− 10.17 ; 3.35) | 0.32 | ||
| Social environment | ||||||
| Parental education levelb,d (ref: Below high-school diploma) | ||||||
| High school to 2-year university degree | 3.50 (− 2.97; 90.98) | 0.29 | 6.09 (− 1.70; 13.89) | 0.13 | ||
| ≥ 3-year university degree | 6.15 (− 1.19; 13.48) | 0.1 | 8.56 (− 0.24; 17.37) | 0.06 | ||
| Number of older siblingsd (ref: no sibling) | ||||||
| 1 | − 3.85 (− 7.58; − 0.12) | 0.04 | − 4.68 (− 9.22; − 0.15) | 0.04 | ||
| ≥ 2 | − 5.10 (− 9.86; − 0.35) | 0.04 | − 7.00 (− 12.80; − 1.21) | 0.02 | ||
| Parental socioeconomic status (SES)c,d (ref: High SES) | ||||||
| Low SES | − 2.84 (− 7.25; 1.58) | 0.21 | − 1.07 (− 6.42; 4.28) | 0.7 | ||
| Middle SES | 0.14 (− 3.83; 4.11) | 0.94 | − 0.95 (− 5.77; 3.86) | 0.7 | ||
| Ever breastfed (yes) | 3.37 (− 0.47; 7.20) | 0.09 | 0.71 (− 3.94; 5.36) | 0.76 | ||
aObese: body mass index ≥ 30 kg/m2.
bParental education: highest level of education of one of the parents.
cLow SES: farm workers, labourers, service workers and unemployed; Medium SES: farmers, supervisors, skilled craftsmen; High SES: business owners, professionals, managers.
dMultiple imputations.
The 31 children (12.8%) with incomplete data had a lower VABS-II daily living score than those with complete data (71.6 (± 11.0) vs. 76.2 (± 13.0), respectively, P = 0.04). Analysis of complete records showed a significant association between breastfeeding and the VABS-II score for daily living skills (adjusted β [95% CI]: 5.15 (1.13; 9.17), P = 0.01) but confirmed the absence of association between breastfeeding and the VABS-II communication score (P = 0.35).
Association between duration of breastfeeding and severity of clinical presentation
In breastfed children, neither the duration of ABF nor the duration of PBF were significantly associated with the severity of clinical presentation for any of the variables considered (VABS-II, SRS-2 T-score, ADOS-2 CSS, IQ and ABC scores; all Spearman's rho < 0.10, with non-significant p-value). Furthermore, no dose effect of ABF or PBF was found on the severity of clinical presentation (see Supplementary Table S3 online).
Discussion
This study examined the relationship of (i) the initiation of breastfeeding and (ii) its duration with the severity of clinical presentation of children diagnosed with ASD according to DSM5 criteria and a multidimensional assessment.
In our sample of 243 children with a confirmed diagnosis of ASD, 73% had been breastfed, which is comparable to the general French population24–26. This contradicts a meta-analysis that found that children with ASD were breastfed less often than the general population suggesting a protective role of breastfeeding against ASD18. However, this result was not confirmed when the diagnosis of ASD had to be made strictly according to DSM criteria (as in our study) and not by parental self-reports. Furthermore, this result was not replicated in a study published by Soke et al.20 (not included in Tseng's meta-analysis) when adjustment was made for confounding factors, as also done in our study.
We did not find any gender difference in our study regarding the prevalence of breastfeeding among children with ASD, which is also in line with the observation in the general population12. We found a 52% rate of children with ASD still being breastfed at six months of age. This is higher than the 19% rate reported in the general paediatric population25, but consistent with the 51% rate reported in another population of children with ASD by Soke et al.20.
In our study, children with ASD who were breastfed had better daily living skills and tended to have better communication skills, according to VABS-II scores, than those who were not. This intergroup difference was not significant after adjusting for a large set of confounding factors related to the child, mother and social environment characteristics. However, our sensitivity analysis carried out only on the children for whom we had all the adjustment variables showed that the initiation of breastfeeding was significantly associated with daily living skills. We assume that this discrepancy was related to a selection bias, the 31 children with missing adjustment variables and who were not included in the sensitivity analysis, having significantly lower daily living skills than those without missing variables.
We did not find any association between the initiation of breastfeeding and the severity of ASD symptoms measured by the ADOS-2 CSS or the SRS-2, which cannot be compared with the literature in the absence of published data. Furthermore, in contrast with the general population9,27,28, our results do not support a protective effect of breastfeeding on the IQ of children with ASD.
Although there is some evidence in previous literature that longer duration of any and predominant breastfeeding are significantly associated with a decreased risk of ASD22, our results do not support a protective role against greater clinical severity in ASD. The evidence of a positive association between breastfeeding duration and IQ in the general population9,10,12,29 was not found in our sample of children with ASD.
This study is one of the first to examine the association between breastfeeding and the severity of clinical presentation in children with ASD. One of the strengths of our study is the large sample of children with a confirmed diagnosis of ASD following a multidisciplinary clinical assessment and using DSM-5 criteria. In addition, the assessment of clinical presentation was performed using standardised tools that target multiple dimensions, such as ASD symptoms, IQ, adaptive behaviours and behavioural problems30. We also considered a wide range of potential confounders, including parental socioeconomic and educational levels16.
Our study has several limitations that need to be taken into account when interpreting the results. First, data on breastfeeding were collected retrospectively when the children were, on average, 10 years old (± 3.86), which implies a risk of recall bias. However, it has been suggested that mothers' recall of breastfeeding initiation and duration are accurate, even 6 to 20 years after childbirth31–34. Second, because participants were recruited from tertiary centres, such as autism resource centres, the study population may not be fully representative of the general population with ASD. Third, the children included in our study differed from the other participants in the ELENA cohort in that they had higher levels of IQ and adaptive skills, which limits the generalisability of our results. Maternal IQ16,35 and autistic traits20, which are also known to be potential confounders, were not taken into account in our study. Finally, regarding the protective effect of breastfeeding, the population size of our study is limited compared to other studies, some involving several thousand participants. It is therefore not possible to exclude a possible association between breastfeeding and severity of clinical presentation in ASD. However, given the limitations of our study and of those published on breastfeeding in ASD, prospective studies of populations at risk of ASD are necessary. These need to be based particularly on large prenatal cohorts, making it possible to take confounding factors into account, including family socio-demographic factors and maternal metabolic conditions, by means of a very large data collection.
Methods
This retrospective survey, carried out between September 20 and November 22, 2021, was nested in the ELENA cohort36, a prospective, multiregional and observational study of children with a confirmed diagnosis of ASD followed for a period of six years. Signed informed consent was obtained from a parent and/or legal guardian for the study participation. The study design and protocol have previously been thoroughly described36. This study was approved by the Review Board of the University Hospital of Montpellier, France. All methods were performed in accordance with the relevant guidelines and regulations.
Participants
Among the 876 children with a confirmed diagnosis of ASD included in the ELENA cohort, 243 (27.7%) parents had fully completed the breastfeeding section in the Early Feeding Patterns Questionnaire. The mean age of the 243 participants was 10.3 years (± 3.9) and the age range was 4 to 22 years. The sample characteristics are detailed in Table 1.
Measurements
Data collection
Parents were sent an email inviting them to participate in the study by completing an online questionnaire on early feeding patterns. Parents who did not complete the questionnaire were contacted again two weeks and one month after the initial request, by email and phone. Only one questionnaire was to be completed per child and by at least one parent. The questionnaire was either self-administered online, administered by phone with the help of a clinical research assistant or completed using a printed version mailed back to our research team.
Early feeding patterns questionnaire
Detailed information was obtained using a parental self-questionnaire adapted from the EDEN cohort37. It focused on (i) parents at the time of the birth: age, education level, socio-economic status (SES), occupation, household composition, mother's weight, height and body mass index before pregnancy, maternal smoking; (ii) birth of the child: gestational age, delivery, birth weight; and (iii) feeding patterns of the newborn in the first year of life: initiation and duration of breastfeeding, respective ages at first formula feeding and solid foods.
Early feeding patterns
Mothers who initiated breastfeeding were identified by the following question: "Did you breastfeed your baby, even for a few hours?”. The duration in months of any breastfeeding (ABF, including partial or exclusive breastfeeding) was then determined by the question: "How old was your baby when you stopped breastfeeding?”. The duration in months of predominant breastfeeding (PBF—where the infant also receives water or other liquids38) was determined by asking the age of initiation of solid foods and/or formula. To facilitate their answer, parents were asked to use the child's health record where this information on breastfeeding and early feeding is usually found.
Child assessment
Severity of the clinical presentation was determined through the assessment of autism symptoms, intellectual level, adaptive behaviours and behavioural problems. These data were collected at the time of inclusion in the ELENA cohort, when the children were on average 5.8 years old (standard deviation (SD) = 3.4). The severity of autism symptoms was measured using the Autism Diagnostic Observation Schedule-2 (ADOS-2) total algorithm score (calibrated severity score, CSS)39,40. The ADOS-2 is a standardised assessment tool based on a set of age-appropriate structured or semi-structured tasks. A total score of 5–7 suggests moderate related symptoms and > 8 high related symptoms. The intellectual quotient (IQ) was calculated using standardised and validated instruments (Brunet-Lézine-R41, BECS42, PEP-343, WPPSI-IV44, WISC-V45, WAIS-IV46, K-ABC47) selected according to the child's age and developmental level following the approach used by Howlin et al.48. Adaptive skills were assessed using the Vineland Adaptive Behavior Scales, Second Edition (VABS-II), a standardised, semi-structured parent interview49. Standard scores (mean = 100, SD = 15) were obtained for the three sub-domains: communication, daily living skills, and socialisation. Deficits in social behaviour were assessed using the Social-Responsiveness Scale, second version (SRS-2), administered to parents50,51. Gender-standardised total T-scores were used. Behavioural problems were measured with the Aberrant Behavior Checklist (ABC) completed by parents52.
Statistical analysis
The main exposures were the initiation of breastfeeding and the durations of ABF and PBF. The outcome was the child’s clinical presentation severity defined by a set of variables (ADOS-2 CSS, VABS-II standard scores in communication, socialisation and daily living skills, IQ, SRS-2 T-total score and ABC scores for irritability/aggressiveness, lethargy/social withdrawal, stereotyped behaviours/self-harm and hyperactivity/lack of cooperation). In our analysis, we took into account multiple confounding factors considered in previous studies on the association between breastfeeding and children's cognitive development16: child's sex, gestational age, mother's age at delivery, maternal obesity or smoking during pregnancy, education and socio-economic status of the household and the presence of siblings.
Continuous variables are described by means ± SDs or medians with interquartile ranges (IQRs) and categorical variables by frequencies and percentages. The Chi-square or Fisher's test and the Wilcoxon Mann–Whitney test or ANOVA were used, as appropriate, to compare differences between children: breastfed or never, included in the study or not, with or without missing data for confounders.
We evaluated associations of (i) the initiation of breastfeeding and (ii) the durations of ABF and PBF among breastfed children on clinical severity outcomes using ANOVA and Spearman’s correlations respectively. When the p-value of the effect was < 0.20, a multiple linear regression analysis adjusted for the confounding variables listed above was computed. The chained equations missing data method53 was used to impute the missing confounder values (15 imputation datasets), assuming that the data were randomly missing. The missing data description is presented in Supplementary Table S1. The goodness of fit was assessed by the R2. A sensitivity analysis was performed to investigate the effect of missing data using a complete case analysis. Among breastfed children, a possible dose–effect of the duration of ABF or PBF on clinical severity outcomes was assessed using ANOVA. ABF and PBF were used in tertiles: low (< 2 months), middle (2–4.5 months) and high (≥ 4.5 months) for PBF duration and low (< 3 months), middle (3–8 months) and high (≥ 8 months) for ABF duration. All statistical tests were considered significant for P < 0.05. Analyses were performed using SAS Enterprise Guide V7.13 (SAS Institute Inc., Cary, NC, USA).
Ethical approval
The study and informed consent procedure have been approved by the Internal Review Board of the University Hospital of Montpellier. The ELENA cohort has been approved by the Ethics Committee on the Research of Human Subjects at Marseille Mediterranean and the National Commission for Computing and Liberties (CNIL number DR-2015–393).
Consent to participate
Signed informed consent has been obtained from all participating families included in the ELENA cohort.
Supplementary Information
Acknowledgements
We warmly thank the contributing families, the ELENA staff (Ela Miniarikova, Lee Audras-Torrent, Noémie Fourgassié, Laura Fragey, Marion Sevens, Julie Loubersac, Cécile Michelon, and Laëtitia Ferrando) and the ELENA study group. A full list of members and their affiliations appears in the Supplementary Information. We also express gratitude to the Caisse Nationale de Solidarité pour l’Autonomie (CNSA) and the French Health Ministry (DGOS) for funding this research.
Author contributions
A.B. is the principal investigator of the ELENA cohort. A.B., M.P. and F.D. contributed to the study conception, design and data collection and analysis. Interpretation of the data was done by A.B., M.P. and J.Y.B. The draft of the manuscript was written by A.B., M.P. and F.D. A.B., M.P., F.D., M.C.P., B.H. and J.Y.B. commented on previous versions of the manuscript, and all authors read and approved the final manuscript.
Funding
This work was supported by the French Health Ministry (DGOS) PHRCN 2013 [Grant number 1: 13-0232] and the Caisse Nationale de Solidarité pour l’Autonomie (CNSA) [Grant number 2: 030319]. The CHU of Montpellier (AOI) provided additional support. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.
Data availability
The datasets used and analysed during the current study are available from the corresponding author upon reasonable request.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-022-27040-x.
References
- 1.Ecker C, Bookheimer SY, Murphy DGM. Neuroimaging in Autism spectrum disorder: Brain structure and function across the lifespan. Lancet Neurol. 2015;14:1121–1134. doi: 10.1016/S1474-4422(15)00050-2. [DOI] [PubMed] [Google Scholar]
- 2.Baghdadli A, et al. Developmental trajectories of adaptive behaviors from early childhood to adolescence in a cohort of 152 children with autism spectrum disorders. J. Autism Dev. Disord. 2012;42:1314–1325. doi: 10.1007/s10803-011-1357-z. [DOI] [PubMed] [Google Scholar]
- 3.Modabbernia A, Velthorst E, Reichenberg A. Environmental risk factors for autism: An evidence-based review of systematic reviews and meta-analyses. Mol. Autism. 2017;8:13. doi: 10.1186/s13229-017-0121-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sandin S, et al. The familial risk of Autism. JAMA. 2014;311:1770–1777. doi: 10.1001/jama.2014.4144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Surén P, et al. Association between maternal use of folic acid supplements and risk of Autism spectrum disorders in children. JAMA. 2013;309:570–577. doi: 10.1001/jama.2012.155925. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Brown AS, van Os J, Driessens C, Hoek HW, Susser ES. Further evidence of relation between prenatal famine and major affective disorder. AJP. 2000;157:190–195. doi: 10.1176/appi.ajp.157.2.190. [DOI] [PubMed] [Google Scholar]
- 7.Brown AS, Susser ES, Lin SP, Neugebauer R, Gorman JM. Increased risk of affective disorders in males after second trimester prenatal exposure to the Dutch Hunger Winter of 1944–1945. Br. J. Psychiatr. 1995;166:601–606. doi: 10.1192/bjp.166.5.601. [DOI] [PubMed] [Google Scholar]
- 8.Jacka FN, et al. Maternal and early postnatal nutrition and mental health of offspring by age 5 years: A prospective cohort study. J. Am. Acad. Child Adolesc. Psychiatr. 2013;52:1038–1047. doi: 10.1016/j.jaac.2013.07.002. [DOI] [PubMed] [Google Scholar]
- 9.Anderson JW, Johnstone BM, Remley DT. Breast-feeding and cognitive development: A meta-analysis. Am. J. Clin. Nutr. 1999;70:525–535. doi: 10.1093/ajcn/70.4.525. [DOI] [PubMed] [Google Scholar]
- 10.Belfort MB, et al. Infant feeding and childhood cognition at ages 3 and 7 years: Effects of breastfeeding duration and exclusivity. JAMA Pediatr. 2013;167:836–844. doi: 10.1001/jamapediatrics.2013.455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Jedrychowski W, et al. Effect of exclusive breastfeeding on the development of children’s cognitive function in the Krakow prospective birth cohort study. Eur. J. Pediatr. 2012;171:151–158. doi: 10.1007/s00431-011-1507-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Bernard JY, et al. Breastfeeding duration and cognitive development at 2 and 3 years of age in the EDEN mother-child cohort. J. Pediatr. 2013;163:36–42.e1. doi: 10.1016/j.jpeds.2012.11.090. [DOI] [PubMed] [Google Scholar]
- 13.Deoni S, Dean D, Joelson S, O’Regan J, Schneider N. Early nutrition influences developmental myelination and cognition in infants and young children. Neuroimage. 2018;178:649–659. doi: 10.1016/j.neuroimage.2017.12.056. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Eickmann SH, et al. Breast feeding and mental and motor development at 12 months in a low-income population in northeast Brazil. Paediatr. Perinat. Epidemiol. 2007;21:129–137. doi: 10.1111/j.1365-3016.2007.00795.x. [DOI] [PubMed] [Google Scholar]
- 15.Drane DL, Logemann JA. A critical evaluation of the evidence on the association between type of infant feeding and cognitive development. Paediatr. Perinat. Epidemiol. 2000;14:349–356. doi: 10.1046/j.1365-3016.2000.00301.x. [DOI] [PubMed] [Google Scholar]
- 16.Walfisch A, Sermer C, Cressman A, Koren G. Breast milk and cognitive development—the role of confounders: A systematic review. BMJ Open. 2013;3:e003259. doi: 10.1136/bmjopen-2013-003259. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ghozy S, et al. Association of breastfeeding status with risk of autism spectrum disorder: A systematic review, dose-response analysis and meta-analysis. Asian J. Psychiatr. 2020;48:101916. doi: 10.1016/j.ajp.2019.101916. [DOI] [PubMed] [Google Scholar]
- 18.Tseng P-T, et al. Maternal breastfeeding and autism spectrum disorder in children: A systematic review and meta-analysis. Null. 2019;22:354–362. doi: 10.1080/1028415X.2017.1388598. [DOI] [PubMed] [Google Scholar]
- 19.Husk JS, Keim SA. Breastfeeding and Autism spectrum disorder in the national survey of children’s health. Epidemiology. 2015;26:451–457. doi: 10.1097/EDE.0000000000000290. [DOI] [PubMed] [Google Scholar]
- 20.Soke GN, et al. Association between breastfeeding initiation and duration and Autism spectrum disorder in preschool children enrolled in the study to explore early development. Autism Res. 2019;12:816–829. doi: 10.1002/aur.2091. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Al-Farsi YM, et al. Effect of suboptimal breast-feeding on occurrence of autism: A case-control study. Nutrition. 2012;28:e27–32. doi: 10.1016/j.nut.2012.01.007. [DOI] [PubMed] [Google Scholar]
- 22.Boucher O, et al. Association between breastfeeding duration and cognitive development, autistic traits and ADHD symptoms: A multicenter study in Spain. Pediatr. Res. 2017;81:434–442. doi: 10.1038/pr.2016.238. [DOI] [PubMed] [Google Scholar]
- 23.Baghdadli A, et al. Adaptive trajectories and early risk factors in the autism spectrum: A 15-year prospective study: Adaptive trajectories and risk factors in autism. Autism Res. 2018;11:1455–1467. doi: 10.1002/aur.2022. [DOI] [PubMed] [Google Scholar]
- 24.Courtois F, et al. Trends in breastfeeding practices and mothers’ experience in the French NutriNet-Santé cohort. Int. Breastfeed. J. 2021;16:50. doi: 10.1186/s13006-021-00397-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Wagner S, et al. Durée de l’allaitement en France selon les caractéristiques des parents et de la naissance. Résultats de l’étude longitudinale française Elfe. Bull. Epidémiologique Hebdomadaire BEH. 2015;2:522–532. [Google Scholar]
- 26.Wagner S, et al. Breastfeeding initiation and duration in France: The importance of intergenerational and previous maternal breastfeeding experiences-results from the nationwide ELFE study. Midwifery. 2019;69:67–75. doi: 10.1016/j.midw.2018.10.020. [DOI] [PubMed] [Google Scholar]
- 27.Horta BL, de Loret MC, Victora CG. Breastfeeding and intelligence: A systematic review and meta-analysis. Acta Paediatrica. 2015;104:14–19. doi: 10.1111/apa.13139. [DOI] [PubMed] [Google Scholar]
- 28.Horta BL, de Sousa BA, de Mola CL. Breastfeeding and neurodevelopmental outcomes. Curr. Opin. Clin. Nutr. Metab. Care. 2018;21:174–178. doi: 10.1097/MCO.0000000000000453. [DOI] [PubMed] [Google Scholar]
- 29.Bernard JY, et al. Breastfeeding, polyunsaturated fatty acid levels in colostrum and child intelligence quotient at Age 5–6 years. J. Pediatr. 2017;183:43–50.e3. doi: 10.1016/j.jpeds.2016.12.039. [DOI] [PubMed] [Google Scholar]
- 30.Weitlauf AS, Gotham KO, Vehorn AC, Warren ZE. Brief report: DSM-5 “Levels of Support:” A comment on discrepant conceptualizations of severity in ASD. J. Autism Dev. Disord. 2014;44:471–476. doi: 10.1007/s10803-013-1882-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Amissah EA, Kancherla V, Ko Y-A, Li R. Validation study of maternal recall on breastfeeding duration 6 years after childbirth. J. Hum. Lancet. 2017;33:390–400. doi: 10.1177/0890334417691506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Cupul-Uicab LA, Gladen BC, Hernández-Ávila M, Longnecker MP. Reliability of reported breastfeeding duration among reproductive-aged women from Mexico. Matern. Child Nutr. 2009;5:125–137. doi: 10.1111/j.1740-8709.2008.00159.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Li R, Scanlon KS, Serdula MK. The validity and reliability of maternal recall of breastfeeding practice. Nutr. Rev. 2005;63:103–110. doi: 10.1111/j.1753-4887.2005.tb00128.x. [DOI] [PubMed] [Google Scholar]
- 34.Natland ST, Andersen LF, Nilsen TIL, Forsmo S, Jacobsen GW. Maternal recall of breastfeeding duration twenty years after delivery. BMC Med. Res. Methodol. 2012;12:179. doi: 10.1186/1471-2288-12-179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Der G, Batty GD, Deary IJ. Effect of breast feeding on intelligence in children: Prospective study, sibling pairs analysis, and meta-analysis. BMJ. 2006;333:945. doi: 10.1136/bmj.38978.699583.55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Baghdadli A, et al. Investigating the natural history and prognostic factors of ASD in children: The multic Entric Longitudinal study of childrEN with ASD-the ELENA study protocol. BMJ Open. 2019;9:e026286. doi: 10.1136/bmjopen-2018-026286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Drouillet P, et al. Maternal fatty acid intake and fetal growth: Evidence for an association in overweight women. The ‘EDEN mother–child’ cohort (study of pre- and early postnatal determinants of the child’s development and health) Br. J. Nutr. 2008;101:583–591. doi: 10.1017/S0007114508025038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.World Health Organization. Indicators for assessing infant and young child feeding practices: Part 1: Definitions: Conclusions of a consensus meeting held 6–8 November 2007 in Washington DC, USA. (World Health Organization, 2008).
- 39.Gotham K, et al. A replication of the Autism diagnostic observation schedule (ADOS) revised algorithms. J. Am. Acad. Child Adolesc. Psychiatr. 2008;47:642–651. doi: 10.1097/CHI.0b013e31816bffb7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Lord, C. et al. Autism diagnostic observation schedule, (ADOS-2) modules 1–4. Los Angeles, California: Western Psychological Services (2012).
- 41.Brunet, O., Lézine, I. & Josse, D. Brunet-Lézine révisé: Echelle de développement psychomoteur de la première enfance: Manuel BLR-C. (Etablissements d’Applications Psychotechniques, 1997).
- 42.Adrien, J.-L. BECS: Pratiques psychologiques et recherches cliniques auprès d’enfants atteints de TED. (De Boeck, 2008).
- 43.Lansing M. D., Marcus L. M., Reichler R. J., & Schopler E. PEP-3 Profil psycho-éducatif. (De Boeck, 2010).
- 44.Wechsler, D. WPPSI-IV, échelle d’intelligence de Wechsler pour enfants. (ECPA, 2014).
- 45.Wechsler, D. WISC-V: Administration and scoring manual. (NCS Pearson, Incorporated, 2014).
- 46.Wechsler D. Wechsler adult intelligence scale–Fourth Edition (WAIS–IV) San Antonio TX NCS Pearson. 2008;22:816–827. [Google Scholar]
- 47.Kaufman, A. S. K-ABC: Kaufman Assessment Battery for Children: Interpretive Manual. (American Guidance Service, 1983).
- 48.Howlin P, Savage S, Moss P, Tempier A, Rutter M. Cognitive and language skills in adults with autism: A 40-year follow-up. J. Child Psychol. Psychiatr. 2014;55:49–58. doi: 10.1111/jcpp.12115. [DOI] [PubMed] [Google Scholar]
- 49.Sparrow, S. S., Cicchetti, D. V., Balla, D. A. & Doll, E. A. Vineland Adaptive Behavior Scales: Survey Forms Manual. (American Guidance Service, 2005).
- 50.Bölte S, Poustka F, Constantino JN. Assessing autistic traits: Cross-cultural validation of the social responsiveness scale (SRS) Autism Res. 2008;1:354–363. doi: 10.1002/aur.49. [DOI] [PubMed] [Google Scholar]
- 51.Constantino JN, et al. Validation of a brief quantitative measure of Autistic traits: Comparison of the social responsiveness scale with the Autism diagnostic interview-revised. J. Autism Dev. Disord. 2003;33:427–433. doi: 10.1023/A:1025014929212. [DOI] [PubMed] [Google Scholar]
- 52.Aman MG, Singh NN, Stewart AW, Field CJ. Psychometric characteristics of the aberrant behavior checklist. Am. J. Ment. Defic. 1985;89:492–502. [PubMed] [Google Scholar]
- 53.White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for practice. Stat. Med. 2011;30:377–399. doi: 10.1002/sim.4067. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and analysed during the current study are available from the corresponding author upon reasonable request.