Research can help inform our parenting decisions, but it’s important to remember its limitations, writes Amy Brown
Recently there has been an influx of parenting guidance urging parents to make ‘data-driven’ parenting decisions. Whilst I’m all for research (and it would be slightly odd for me as an academic not to be), it’s important to be able to take a step back, recognise that science is not perfect, and ask how it can help us best when it comes to any decision. Here are three top things to remember:
1. A lack of evidence is not the same as evidence of harm
A lack of research evidence is not proof that something is not important. Sometimes research for the most obvious things doesn’t get conducted, because, well… it’s obvious. The Christmas letter in the BMJ in 2003 summed this up well. Entitled ‘Parachute use to prevent death and major trauma related to gravitational challenge: systematic review of randomised controlled trials’, it concluded that because there were no randomised controlled trials of whether using a parachute when jumping out of a plane had any benefit, there was no evidence that parachutes were needed when jumping out of a plane.
Satire of course, but with a very important point. We don’t need research to justify caring responsively for our babies. Do we really need research that examines whether leaving babies to cry for hours is harmful? Or might the money be better invested in understanding how to support new parents?
It’s also important to remember that gaps in our research evidence are in part a consequence of a patriarchal culture. Research that benefits rich, white men still gets the majority of the funding, meaning research into babies and parenting (seen as ‘women’s issues’) often gets left behind.
Just because science hasn’t caught up with what women have known for generations, it doesn’t mean those behaviours and rituals should be dismissed. And as always, the burden of proof should fall on any intervention to prove its safety and benefit, not for biologically normal behaviours to prove their worth. Benefits of breathing, anyone? Exactly.
2. Research methods that are considered ‘gold standard’ often aren’t very helpful when it comes to decisions around motherhood
Many people will tell you that the ‘best’ (or sometimes only) type of research you should listen to is the randomised controlled trial. This works by randomising participants into different groups and then varying the experience of those groups to see if any difference in outcomes occurs. For example, one group might take one medication, another group a different medication, and their outcomes would be compared.
The theory behind trials is that if you genuinely randomise enough people to a group the groups should be pretty equal in terms of everything else apart from whatever you are testing. So other things like age, income, and so on should naturally end up similar in each group. This reduces any factors that might be associated with someone choosing to follow in real life a behaviour that might be associated with a particular outcome.
Randomised controlled trials have their place, but they were never meant to be the only type of worthwhile research, especially outside of clinical medicine. And they are tricky to do well when it comes to human behaviour, because researchers would not ethically be allowed to randomise some behaviours – such as telling women whether to breastfeed or not, since that can do harm. In which case, why do we think we need more research evidence on these things in the first place?
Even if you can randomise a behaviour, people often don’t stick to what you have asked them to do, because lives are complicated. Even if you could randomise one group of women to breastfeed, we know there are so many barriers to them doing so that it would be unlikely they all would. So adherence rates are often low. In one trial asking parents to introduce solids at 3 months or at 6 months, just 43% in the 3 months group fully adhered to instructions, with the biggest reason for not doing so being ‘Maternal distress’.1
A big problem with this is that if you then compare outcomes for the two groups, what are you comparing? And if you only compare people who stuck to the instructions, your groups are no longer random, as other factors will affect who was able to adhere. This matters. In one trial, breastfed babies who had lost a lot of weight were randomised either to continue to exclusively breastfeed, or to have small amounts of formula. The researchers then looked at the microbiome of the babies and declared that this early use of formula had no impact on the microbiome at 1 month, as there were no differences between the two groups. However, by this point babies in the exclusively breastfeeding group were actually having more formula overall per day than the early supplementation group, making the groups quite similar.2
3. Sometimes stories about research miss out important details
Headlines often differ from the reality of what was published. One great example of this is a well-publicised study that looked at the outcomes of controlled crying. The researchers randomised mothers who stated their baby had a sleep problem (based on maternal perception) to receive a controlled crying intervention, or not. They measured maternal mental health and perceptions of infant sleep before and after the intervention. The findings showed that at 10 months and 12 months the controlled crying group reported fewer sleep difficulties than the control group, with a lower incidence of postnatal depression at 10 months.3
These findings are often used to tell parents that controlled crying helps babies. But there is a big error in this message that is clear to see when you properly dig into the research paper. Parents in the controlled crying group were also given information about normal sleep patterns, an opportunity to talk to a nurse about their baby’s sleep, and advice on how to look after their own wellbeing.
It is more likely that this wider support helped them determine whether their baby actually had a sleep problem, and the wider care they received helped them feel that someone was listening – especially compared to the control group, who declared they had a problem but were given no help. When asked which elements of the intervention mothers found helpful, the top agreement at 93% was ‘Having someone to talk to’. This goes to show just how important understanding the full picture is when reporting whether a study is significant or not.
Overall, science can be helpful, but it’s important we recognise its limitations and that it is not the be all and end all. Your instincts and desires matter too. You do not need to present a barrage of evidence for every parenting decision you make. And you certainly do not need permission from science to make only the decisions it has deemed worthy.
Amy Brown is Professor of Child Public Health at Swansea University. She is passionate about understanding how parents can best be supported in feeding their babies, and about helping communicate how research can help inform but not direct the decisions they make. Her fourth book, Informed is Best: Your Straight-talking, Evidence-based Guide to Making the Right Decisions for your Pregnancy, Birth and Baby, is published by Pinter & Martin.
- Michael R. Perkin et al., ‘Enquiring About Tolerance (EAT) study: feasibility of an early allergenic food introduction regimen’, The Journal of Allergy and Clinical Immunology, 137:5 (May 2016), 1477–86.
- Valerie J. Flaherman et al., ‘The effect of early limited formula on breastfeeding, readmission, and intestinal microbiota: a randomized clinical trial’, Journal of Pediatrics, 196 (May 2018), 84–90.
- Harriet Hiscock et al., ‘Improving infant sleep and maternal mental health: a cluster randomised trial’, Archives of Disease in Childhood, 92:11 (November 2007), 952–8.
First published in issue 63 of JUNO. Accurate at the time this issue went to print.