What Four Years of Research Taught Me About Treating Depression In Pregnancy
In November 2022, I took my first step into science as a full-time researcher. I had no idea then that the very first project I worked on would take nearly four years to reach publication. In this article, I will talk about what truly works to treat depression in pregnancy, and why I do this research.
The What and the Why
Depression in pregnancy, which is known as antenatal depression (AD), affects nearly 30% of pregnant women around the world.
What makes AD so important to treat is the negative impacts it has on both the mum as well as her baby. These include the increased risk of preterm birth (babies born before 37 weeks of pregnancy are completed), babies having low birthweight, and additional concerns that persist well beyond infancy.
For mums, antenatal depression remains one of the strongest risk factors for postnatal depression, or depression after birth.
Postnatal depression is widely recognised and receives much of the attention in the public domain. However, it is just as important to take a step back and examine mental health during pregnancy, not just after birth.
Despite emerging research showing that AD can have a profound and long-lasting impact for both mum and baby, it isn’t talked about enough. And this is the remit of the HappyMums consortium, which you can learn more about in this ITM piece. The research I will talk about today was done as part of this large consortium, focusing entirely on mental health during pregnancy.
The How
We wanted to look at the different types of treatment pathways available to treat depression in pregnancy. Our aim was to bring together the entire evidence base of interventions and quantitatively evaluate the effectiveness of each.
So, we used a statistical method called a meta-analysis, which analyses and combines the overall effect sizes (quantitative estimates of how effective a treatment is for a condition), across numerous research studies.
This included psychological interventions like Cognitive Behaviour Therapy (CBT), counselling, and mindfulness for example, and also non-psychological interventions like Omega-3 supplementation, yoga, and music therapy, to name a few.
We decided to focus on randomised controlled trials, the ‘gold-standard’ research methodology to evaluate interventions. In a randomised controlled trial, participants are randomly allocated to either the treatment or a control group to evaluate the true effectiveness.
The What Did We Find?
As we were looking at any peer-reviewed (a quality control process where experts in a specific field evaluate research, scholarly work, or professional performance before it is approved for publication) studies ever published, we had over 2,000 potential studies to filter.
Overall, in our analysis, we included 115 studies from all over the world, with over 12,561 participants, and a wide range of interventions. We found that all the available interventions, regardless of type or format, were effective in treating depression.
Of course, given that the interventions were so varied in duration, method of delivery, and the way the outcome was measured, the studies were extremely heterogeneous. So, after accounting for any ‘outlying’ studies, we ran another analysis, which still found a modest treatment effect.
This made me curious to dig deeper, and to see whether any specific intervention formats were more effective than others. For example, were interventions delivered face-to-face more effective than those which were digitally delivered, and were psychological interventions more effective than others?
Short answer: no.
We did what is called a subgroup analysis, which meant splitting the research data into smaller subsets, like intervention type, format, and location of delivery.
Very interestingly, we did not find any significant differences between intervention type and format. This meant that non-psychological interventions were just as effective as psychological interventions. Furthermore, interventions delivered digitally (through a tablet or smartphone application) were as effective as those delivered face to face.
This does not mean that non-psychological interventions should take precedence over treatments like CBT, which have decades of research and proven efficacy. This means that we have more options than we originally thought in treating AD. We have previously seen, from the SHAPER Melodies for Mums trial, how singing has proven to be effective in treating postnatal depression.
Our findings do not mean that these interventions should replace current guidelines. In the UK, the official treatment guidelines for moderate to severe depression recommend antidepressant medication and high-intensity psychological intervention. And I agree, it should remain this way.
What our do findings show is that when health services face such high demand, leading to long waiting lists, these adjunctive therapies can, and have been shown to be effective. This allows for more flexibility and for earlier treatment, which, in this critical period, is so important.
Talking about intervention formats, I was pleasantly surprised to see that digital interventions are as effective as those delivered in-person. This allows for so much flexibility, especially if a pregnant woman faces barriers related to physical health or distance.
We also found that ‘combination’ interventions, integrating these different aspects, for example psychological therapy with aerobics, and in person interventions with digital support between sessions, were most effective. We proposed that having a flexible approach also allows for intervention facilitators to engage with participants in-between physically delivered sessions and may have higher levels of participant engagement.
Finally, we wanted to see whether biological improvements also align with the psychological improvements. We did another sub-analysis of studies which reported on biological components like cortisol (our primary stress hormone), cholesterol (a marker related to heart health), and other inflammation-associated components. What we found is a clear ‘stress reduction’ effect associated with these biological components.
So, our findings show that these interventions not only improve physical health, but also the underlying biological mechanisms.
The ‘What Next’?
Our study, to our knowledge, is the most comprehensive meta-analysis which demonstrates the effectiveness of interventions in the management of AD. Hopefully, our review shows that there are more interventions that have proven effectiveness than just one, giving more flexibility in treating women earlier on.
What particularly excited me was the finding that digitally delivered interventions show comparable efficacy. And this gets me thinking: if the treatment of AD can be digitalised, could the digital screening of women at risk of developing depression follow a similar pathway?
And spoiler alert, this is exactly what we’re investigating in the HappyMums Smartphone App study.
Stay tuned!
