select search filters
briefings
roundups & rapid reactions
Fiona fox's blog
A case-control study published in PLOS One looks at timely vocal responses in mother-infant interactions and later psychiatric diagnoses.
Dr Catalina Suarez-Rivera, Reader of Child Psychology, University of East London, said:
“The press release accurately reflects the study’s methods, findings, and implications, with one exception. The opening statement—“When moms were slower to vocally respond to their 1-year-old baby’s vocalizations, the child was more likely to have been diagnosed with a psychiatric disorder by age 7”—is not an accurate description of the analysis. The study did not examine the promptness of individual responses in relation to later psychiatric diagnoses, but rather the likelihood that mothers responded within a 1-second threshold. The statement should therefore be framed in terms of maternal responsiveness relative to this 1-second threshold.
“Overall, this is good-quality research, and the conclusions are supported by solid data. I do not have concerns that the authors failed to account for important confounding factors.
“The study has several limitations, all of which are acknowledged by the authors. In my view, the most important are the small sample sizes for some psychiatric disorders, the fact that the quality of maternal responses was not examined, and the limited ability to identify true responses to preceding vocalizations because of poor video quality. The press release mentions the first two limitations but not the third, which I consider important.
“If I were reviewing this paper, I would raise three points, although I do not believe they undermine the validity of the main findings or their implications. First, the authors report 90% confidence intervals for the linear regression models, whereas 95% confidence intervals are used elsewhere in the paper. No justification is provided, and the lower confidence level 90% corresponds to a less stringent significant threshold so effects that would not be statistically significant (or confidence intervals that would cross 0) at the conventional 5% level may appear significant. Second, it is unclear how the final sample of 158 mother–child dyads were selected from the 1,240 children who attended the Child in Focus clinics and completed the book-reading task. Given the large reduction in sample size, the inclusion criteria should be clearly explained to rule out concerns about selection bias. Third, in Table 3, statistical significance should be indicated by confidence intervals that do not cross 1, rather than 0, because the analyses report odds ratios and 1 is the appropriate null value.
“Overall, this work fits well within the existing literature and advances our understanding of how features of early caregiver–child interactions relate to later development. It also makes an important contribution by identifying potential early markers of child psychopathology.”
Dr Marina Bazhydai, Lecturer in Developmental Psychology, Lancaster University, said:
“I found the press release to accurately reflect the scientific contribution of the target paper. The published paper is based on a theoretically driven research hypothesis which is addressed using appropriate methodology and statistical analyses. The authors, as well as the press release, acknowledge the limitations of this research and caution against attributing generalisability to these findings, effectively guarding against overspeculation. The implications for real world include strongly acknowledging the role of primary caregivers on infant development and social, cognitive, and emotional flourishing”.
“The key limitations of this research include the relatively small sample size of 55 cases in which children went on to develop a psychiatric disorder, and the unevenly distributed and limited types of diagnoses captured (only 6 cases of autism). The limitations were duly acknowledged by the authors. This might explain why some of the findings are inconsistent with hypotheses derived from prior literature. Further, the analyses did not account for quality of parent-child interaction, solely focussing on probability of contingent responses, and may lack ecological validity overall due to the lab-based nature of the study (as opposed to natural interactions parents and children might have at home). Finally, while maternal depression was controlled for, other maternal mental health difficulties were not (ADHD, autism).”
Prof Asma Khalil, Professor of Obstetrics and Maternal Medicine and Consultant Obstetrician, City St George’s, University of London (CSGUL), said:
“This is an interesting study that takes advantage of a well-established longitudinal birth cohort and prospectively examines whether aspects of early mother-infant interaction are associated with later childhood psychiatric diagnoses. One of its strengths is that the interaction was assessed years before the diagnosis, reducing the risk of recall bias.
“However, this is an observational study and cannot establish that slower maternal responses cause later behavioural disorders. Parent-child interactions are complex and bidirectional, and the way a parent responds may itself be influenced by the child’s early temperament, neurodevelopmental characteristics or shared genetic factors. Although the authors adjusted for some potential confounders, residual confounding cannot be excluded.
“Another important limitation is the relatively small sample size, particularly within individual diagnostic groups, which means some estimates are imprecise and require confirmation in larger studies.
“Overall, the findings add to growing evidence that early parent-child interactions may provide useful insights into child development, but they should not be interpreted as evidence that delayed maternal responses lead to psychiatric disorders, nor should they be used to blame parents. Further research is needed before these observations can be considered for clinical screening or intervention.”
Prof Swaran Singh, Professor of Social and Community Psychiatry, University of Warwick, said:
“This is an interesting, carefully cautious study. It suggests that early timing of maternal vocal response may carry information about later vulnerability to disruptive behaviour disorders and ADHD.
“But timing is not causation. Interaction is bidirectional; infant temperament, maternal mental health, social adversity, genetic risk, and the wider caregiving environment all shape what is observed. What counts as a “slow” response also varies by culture, there is unlikely to be a single ‘tempo’ for maternal response. The study is small, particularly in the diagnostic subgroups, and it captures timing alone, not the warmth or context of the interaction. A timing measure without emotional content is a limited signal and should not be mistaken for conclusive proof.
“The study should not make mothers feel blamed or anxious. Such disorders carry a long, unfair history of parental blame, and this field has worked hard to move past it. The contribution of this study is real but modest; one more piece in an early literature suggesting interactional patterns may eventually help identify families for supportive, non-stigmatising intervention, but only once replicated, and only ever alongside other markers, never alone. For now, this is hypothesis-generating, not a basis for screening or predicting outcomes in individual children.”
Prof Alastair Sutcliffe, Professor of General Paediatrics, University College London (UCL), said:
“This is an interesting exploratory study using prospectively collected ALSPAC data, but it does not show that slower maternal responses cause psychiatric disorders or can yet reliably predict them.
“The analysis included only 158 mother–infant pairs, including just 16 children categorised as having ADHD, and the one-second response threshold was selected after examining the same dataset. This increases the possibility that the findings are optimistic or due partly to chance.
“The adjustment for confounding was limited, so shared genetic factors, parental neurodevelopmental traits, family circumstances or early characteristics of the child could explain the association.
“The recordings were brief, clinic-based and of limited sound quality, and measured only the timing—not the content, warmth or appropriateness—of the mother’s response. This is a key limitation.
“The authors describing the association as ‘robust’ and discussing screening is premature because the study provides no sensitivity, specificity or independent validation.”
“Parents should certainly not conclude that pausing before responding to their baby causes ADHD or behavioural problems.”
Prof Dennis Ougrin, Professor of Child and Adolescent Psychiatry and Global Mental Health at Queen Mary said:
“The study has several strong points. It is based on objective measures of mother-child interactions, rather than self-reported measures. It is based on a large and very well-characterised cohort. It also builds on many other studies showing that attunement between parents and children is important for child development.
“However, I would caution against interpreting these findings as causal. It may well be the case that some other factors, such as genetic factors, may have influenced both the speed of maternal response and the risk of mental health disorders in children. Nevertheless, the findings represent an interesting and novel risk factor which could generate new research hypotheses.”
‘Probability of a timely vocal response in mother-infant interaction and later psychiatric diagnosis: A case-control study’ by Bethany Stanley et al. was published in PLOS One at 19:00 UK time on Wednesday the 1st of July 2026.
DOI: https://doi.org/10.1371/journal.pone.0344552
Declared interests
Dr Marina Bazhydai: “I declare no conflict of interest.”
Prof Asma Khalil: “No relevant COI”
Prof Swaran Singh: “No Conflicts of Interest declared”
Prof Alastair Sutcliffe: “no conflict of interest “
Prof Dennis Ougrin: “None.”