Long-term behavioural consequences of infant feeding: The limits of observational studies

Michael S. Kramer, Eric Fombonne, Lidia Matush, Natalia Bogdanovich, Mourad Dahhou, Robert W. Platt

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Observational (non-experimental) studies of the association between infant feeding and subsequent child or adult behaviour are prone to residual confounding by subtle differences in psychological attributes and interactional styles of mothers who breast feed vs. those who formula-feed. We followed up 13 889 6.5-year-old Belarusian children who participated in a large cluster-randomised trial of a breast-feeding promotion intervention. Behaviour was evaluated using the Strengths and Difficulties Questionnaire (SDQ), completed independently by the children's parents and teachers. We compared the results of experimental (intention-to-treat, ITT) and observational analyses (based on feeding actually received), both adjusted for clustering. Observational analyses were additionally adjusted for geographical region, urban vs. rural residence, child's sex, age at follow-up, birthweight, and maternal and paternal education. No differences between the randomised experimental vs. control groups were observed in ITT analyses. In contrast, small but statistically significant associations with weaning prior to 3 months were observed for parent and teacher SDQ scores on total difficulties, conduct problems and hyperactivity, even after multivariable adjustment. The absence of associations based on ITT analyses, in contrast with the significant associations based on observed breast-feeding duration, strongly suggests that the latter are biased by residual confounding.

Original languageEnglish (US)
Pages (from-to)500-506
Number of pages7
JournalPaediatric and Perinatal Epidemiology
Volume25
Issue number6
DOIs
StatePublished - Nov 2011
Externally publishedYes

Fingerprint

Intention to Treat Analysis
Observational Studies
Breast Feeding
Mothers
Social Adjustment
Weaning
Cluster Analysis
Breast
Parents
Psychology
Education
Control Groups
Surveys and Questionnaires

Keywords

  • breast feeding
  • child behaviour
  • RCT
  • weaning

ASJC Scopus subject areas

  • Epidemiology
  • Pediatrics, Perinatology, and Child Health

Cite this

Long-term behavioural consequences of infant feeding : The limits of observational studies. / Kramer, Michael S.; Fombonne, Eric; Matush, Lidia; Bogdanovich, Natalia; Dahhou, Mourad; Platt, Robert W.

In: Paediatric and Perinatal Epidemiology, Vol. 25, No. 6, 11.2011, p. 500-506.

Research output: Contribution to journalArticle

Kramer, Michael S. ; Fombonne, Eric ; Matush, Lidia ; Bogdanovich, Natalia ; Dahhou, Mourad ; Platt, Robert W. / Long-term behavioural consequences of infant feeding : The limits of observational studies. In: Paediatric and Perinatal Epidemiology. 2011 ; Vol. 25, No. 6. pp. 500-506.
@article{220c2cb64a9945ddbf1642092ea2de38,
title = "Long-term behavioural consequences of infant feeding: The limits of observational studies",
abstract = "Observational (non-experimental) studies of the association between infant feeding and subsequent child or adult behaviour are prone to residual confounding by subtle differences in psychological attributes and interactional styles of mothers who breast feed vs. those who formula-feed. We followed up 13 889 6.5-year-old Belarusian children who participated in a large cluster-randomised trial of a breast-feeding promotion intervention. Behaviour was evaluated using the Strengths and Difficulties Questionnaire (SDQ), completed independently by the children's parents and teachers. We compared the results of experimental (intention-to-treat, ITT) and observational analyses (based on feeding actually received), both adjusted for clustering. Observational analyses were additionally adjusted for geographical region, urban vs. rural residence, child's sex, age at follow-up, birthweight, and maternal and paternal education. No differences between the randomised experimental vs. control groups were observed in ITT analyses. In contrast, small but statistically significant associations with weaning prior to 3 months were observed for parent and teacher SDQ scores on total difficulties, conduct problems and hyperactivity, even after multivariable adjustment. The absence of associations based on ITT analyses, in contrast with the significant associations based on observed breast-feeding duration, strongly suggests that the latter are biased by residual confounding.",
keywords = "breast feeding, child behaviour, RCT, weaning",
author = "Kramer, {Michael S.} and Eric Fombonne and Lidia Matush and Natalia Bogdanovich and Mourad Dahhou and Platt, {Robert W.}",
year = "2011",
month = "11",
doi = "10.1111/j.1365-3016.2011.01211.x",
language = "English (US)",
volume = "25",
pages = "500--506",
journal = "Paediatric and Perinatal Epidemiology",
issn = "0269-5022",
publisher = "Wiley-Blackwell",
number = "6",

}

TY - JOUR

T1 - Long-term behavioural consequences of infant feeding

T2 - The limits of observational studies

AU - Kramer, Michael S.

AU - Fombonne, Eric

AU - Matush, Lidia

AU - Bogdanovich, Natalia

AU - Dahhou, Mourad

AU - Platt, Robert W.

PY - 2011/11

Y1 - 2011/11

N2 - Observational (non-experimental) studies of the association between infant feeding and subsequent child or adult behaviour are prone to residual confounding by subtle differences in psychological attributes and interactional styles of mothers who breast feed vs. those who formula-feed. We followed up 13 889 6.5-year-old Belarusian children who participated in a large cluster-randomised trial of a breast-feeding promotion intervention. Behaviour was evaluated using the Strengths and Difficulties Questionnaire (SDQ), completed independently by the children's parents and teachers. We compared the results of experimental (intention-to-treat, ITT) and observational analyses (based on feeding actually received), both adjusted for clustering. Observational analyses were additionally adjusted for geographical region, urban vs. rural residence, child's sex, age at follow-up, birthweight, and maternal and paternal education. No differences between the randomised experimental vs. control groups were observed in ITT analyses. In contrast, small but statistically significant associations with weaning prior to 3 months were observed for parent and teacher SDQ scores on total difficulties, conduct problems and hyperactivity, even after multivariable adjustment. The absence of associations based on ITT analyses, in contrast with the significant associations based on observed breast-feeding duration, strongly suggests that the latter are biased by residual confounding.

AB - Observational (non-experimental) studies of the association between infant feeding and subsequent child or adult behaviour are prone to residual confounding by subtle differences in psychological attributes and interactional styles of mothers who breast feed vs. those who formula-feed. We followed up 13 889 6.5-year-old Belarusian children who participated in a large cluster-randomised trial of a breast-feeding promotion intervention. Behaviour was evaluated using the Strengths and Difficulties Questionnaire (SDQ), completed independently by the children's parents and teachers. We compared the results of experimental (intention-to-treat, ITT) and observational analyses (based on feeding actually received), both adjusted for clustering. Observational analyses were additionally adjusted for geographical region, urban vs. rural residence, child's sex, age at follow-up, birthweight, and maternal and paternal education. No differences between the randomised experimental vs. control groups were observed in ITT analyses. In contrast, small but statistically significant associations with weaning prior to 3 months were observed for parent and teacher SDQ scores on total difficulties, conduct problems and hyperactivity, even after multivariable adjustment. The absence of associations based on ITT analyses, in contrast with the significant associations based on observed breast-feeding duration, strongly suggests that the latter are biased by residual confounding.

KW - breast feeding

KW - child behaviour

KW - RCT

KW - weaning

UR - http://www.scopus.com/inward/record.url?scp=80054087234&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80054087234&partnerID=8YFLogxK

U2 - 10.1111/j.1365-3016.2011.01211.x

DO - 10.1111/j.1365-3016.2011.01211.x

M3 - Article

C2 - 21980939

AN - SCOPUS:80054087234

VL - 25

SP - 500

EP - 506

JO - Paediatric and Perinatal Epidemiology

JF - Paediatric and Perinatal Epidemiology

SN - 0269-5022

IS - 6

ER -