Skip navigation

Infant massage: understanding the evidence base


28 Sep 2015

What do we know and not know about the evidence underpinning infant massage, and what are the implications for providers and commissioners?

As a What Works Centre, a key part of the Early Intervention Foundation’s (EIF) role is to evaluate the strength of evidence underpinning programmes and practices across the country. This knowledge is vital of improving child outcomes and using public resources effectively.

In a recent Guardian piece — Evidence doesn’t have a to be a straitjacket for social care practice — Dez Holmes, Director of Research in Practice, and Donna Molloy, EIF’s Director of Implementation, observe that many popular practices such as infant massage have relatively little evidence to support their use.

The practice

Infant massage is commonly practiced in non-western cultures to promote infant sleep and aid digestion.[1,2] The practice was introduced in western cultures approximately 30 years ago through neonatal wards to support the development of premature infants residing in intensive care units.[3] Infant massage has since been implemented more widely to support maternal sensitivity and potentially reduce symptoms of maternal depression.[4] Its widespread use is based on the assumption that the types of touch facilitated through infant massage could also increase mothers’ awareness of their infants’ cues and support the early attachment relationship.[5]

Infant massage can be taught to parents individually, but is more often delivered to small groups of mothers who attend weekly classes when their infant is between six and eight weeks old. Group-based programmes are typically led by a single practitioner who demonstrates massage techniques on a doll. Parents then practice the techniques with their infant as they receive coaching from the practitioner. Sessions usually last an hour and are offered for a period of four to six weeks. The practice can be delivered by a variety of practitioners with various professional qualifications.

The evidence

Infant massage has undergone multiple evaluations, although relatively few have been conducted to a high enough standard to confidently link the practice to any positive parent or child outcomes. Those that have suggest that the benefits of the activity are largely unknown and potentially limited. For example, a recent Cochrane review found that only 14 out of 34 randomised controlled trials involving infant massage were sufficiently rigorous to assess its impact.[6] These studies observed that the benefits of infant massage are either weak or short lived, leading the authors to conclude that research to date does ‘not currently support the use of infant massage with low-risk groups of parents and infants’. Additional evidence suggests that infant massage could potentially make things worse for mothers and infants whose relationship was observed to be ‘at risk’.[7]

Two small-scale RCTs have found, however, that infant massage may be beneficial for mothers at risk of postnatal depression. Specifically, these studies observed that mothers attending an infant massage course were significantly less likely to report clinical symptoms of depression than were mothers not attending infant massage. However, the EIF evidence team has examined these studies in greater depth and found their conclusions to be questionable because of significant problems with their evaluation design.  Specifically:

  • The first study[8] involved a small sample of mothers (34) identified as at risk of depression. Seven out of the 19 assigned to infant massage dropped out of programme and the study. The 12 who remained were not actually depressed at the beginning of the programme, meaning that they were not representative of depressed mothers more generally.
  • The second study[9] involved a larger sample of mothers (62; 31 attending infant massage and 31 attending a support group). However, this study only observed a statistically significant difference in the percentage of mothers scoring in the clinical range for depression at the end of the programme, not a significant difference in the scores between the two groups. Moreover, there was no difference between the groups in terms of child outcomes, and for one outcome, the support group outscored the infant massage group.

A third evaluation examining the processes and outcomes of infant massage additionally observed no specific benefits for low risk mothers, as well as a potential increase in problems for high risk mothers.[10] While the study observed increases in the sensitivity of mothers at moderate risk, the findings are speculative because 1) they involved an extremely small sample of only five mothers and 2) there was no comparison group. The study went on to identify programme and participant characteristics that could potentially improve the benefits of infant massage programmes, but these characteristics have yet to be carefully evaluated.

Thus, the evidence linking infant massage to improvements in maternal mood is not particularly strong or convincing. This evidence should also be considered in light of findings from other, more rigorous studies, suggesting that maternal sensitivity is difficult to improve when the mother is depressed. Specifically, these studies recommend that more intensive interventions, lasting over a period of months or even a year, may be necessary to increase maternal sensitivity and support the attachment relationship when the mother is depressed.[11,12,13,14] From this perspective, infant massage techniques, on their own, are unlikely to be sufficient to increase depressed or otherwise at-risk mothers’ ability to respond sensitively to their child.

In sum, the evidence to date suggests that while the practice of infant massage is popular and is likely not to be harmful, there is no conclusive evidence to suggest that it provides lasting benefits for low or high risk mothers and their infants, including mothers who are feeling depressed.

Implications for commissioning

The example of infant massage highlights how many common and popular practices have relatively little evidence to support their widespread use. Does this then suggest that non-evidence-based practices, such as infant massage, be decommissioned?

The answer to this question is — it depends. The decision to decommission a popular practice with relatively little evidence should be based on a variety of factors, including its cost (a theme of much recent work at the EIF) and the reasons why it is being commissioned. If the activity is safe, relatively inexpensive and commissioned primarily because local families want and enjoy it — than there is likely no harm in making it available to parents and children, or for communities to commission it for themselves. However, if the activity is being offered to improve outcomes that it has little evidence of achieving – then commissioners should think twice about why and for whom they are commissioning it, particularly when resources are very tight.

This is especially true for interventions and practices offered to children and parents who are at risk of various adversities, including mental health problems. In these cases, activities like infant massage may not directly harm families – but they could indirectly make things worse for them by interfering with their access to more effective interventions. On the basis of the evidence, infant massage might therefore be offered to mothers and infants because they want and enjoy it, but should not be commissioned to support mental health outcomes that it has little evidence of achieving.

From this perspective, a lack of evidence does not mean a programme is not of value. But it does mean that commissioners and practitioners should have a clear understanding about what this value is and whether it is sufficient for the outcomes they aim to improve.


[1] Field T., Schanberg S., Davalos M., Malphurs J. (1996) Massage with oil has more positive effects on normal infants. Journal of Prenatal and Perinatal Psychology and Health, 11, 75–80.

[2] Field T. (2000) Infant Massage Therapy. Ch. 32, In: Zeanah C. (Ed) Handbook of Infant Mental Health. Second Edition. New York: The Guilford Press.

[3] Bennett, C., Underdown, A. and Barlow, J. (20103). Massage for promoting mental and physical health in typically developing infants under the age of six months (Review) Issue 4, The Cochrane Library:  John wiley and Sons.

[4] Peláez-Noguera, M., Field, T.MN., Hossain, Z. and Picken, J. (1996). Depressed Mothers’ Touching Increases Infants’ Positive Affect and Attention in Still-Face Interactions. Child Development, 79, 67, 1780–1792.

[5] Underdown, A., Barlow, J., (2011). Interventions to support early relationships:  Mechanisms identified within infant massage programmes. Community Practitioner, 84, 21–26.

[6] Bennett, C., Underdown, A. and Barlow, J. (20103). Massage for promoting mental and physical health in typically developing infants under the age of six months (Review) Issue 4, The Cochrane Library:  John wiley and Sons.

[7] Underdown A., Norwood R., Barlow J. (2013) A Realist Evaluation of the Processes and Outcomes of Infant Massage Programs. Infant Mental Health Journal, 34 (6):483–495.

[8] Onozawa, K., Glover, V., Adams, D., Modi, N., and Kumar, R.C. (2001).  Infant massage improves mother-infant interaction for mothers with postnatal depression. Journal of Affective Disorders, 63, 201–207.

[9] O’Higgins, M., St. James Roberts, I., and Glover, V. (2008).  Postnatal depression and mother infant outcomes after infant massage.  Journal of Affective Disorders. 109, 189–192.

[10] Underdown A., Norwood R., Barlow J. (2013) A Realist Evaluation of the Processes and Outcomes of Infant Massage Programs. Infant Mental Health Journal. 34, 483–495.

[11] Cooper, P.J., De Pascalis, L., Woolgar, M.,Romaniuk, H. and Murray, L. (2015). Attempting to prevent postnatal depression by targeting the mother-infant relationship:  A randomised controlled trial. Primary Health Care Research and Development, 16, 383–397.

[12] Murray, L., Cooper, P., Arteche, A., Stein, Al, and Tomlinson, M. (2015). Randomized controlled trial of a home-visiting intervention on infant cognitive development in per-urban South Africa, DOI: 10.1111/dmcn.12873.

[13] Cooper, P.J., Tomlinson, M., Swartz, L., Landman, M., Monteno, C. Stein, A., McPherson, K., and Murray, L. (2009). Improving quality of mother-infant relationship and infant attachment in socioeconomically deprived community in South Africa: randomised controlled trial, BMJ, 338, 1–8.

[14] Cicchetti, D., Toth, S.L. and Rogosch, F.A. (1999). The efficacy of toddler-parent psychotherapy to increase attachment security in off-spring of depressed, mothers. Attachment and Human Development, 1, 34–66.

About the author

Dr Kirsten Asmussen

Kirsten is head of what works, child development, at EIF.