Body weight perinatal depression: what's the link?

Mature,Brunette,Fat,Woman,In,Green,Garden,Wearing,Hat,,Smiling,

Antenatal and postnatal depression affect around 1 in 8 and 1 in 10 women, respectively (NHS, 2021). Both conditions are understandably difficult for women, causing them to feel sad, guilty, and uninterested at a time when they could be buying booties, morphing into DIY experts to babyproof the flat, and forming an impenetrable bond with the life they brought into the world (no pressure though!). There can also be long-term consequences for their children, including emotional, developmental, and academic problems (Dadi et al, 2020; Slomian et al, 2019).

Research suggests that women who are overweight or obese when they become pregnant are more at risk of suffering from antenatal and postnatal depressive symptoms (e.g. Molyneaux et al, 2014). With roughly 12.3% of women entering pregnancy overweight or obese (Zhao et al, 2018), it may be unsurprising that antenatal and postnatal depressive symptoms are so common. But let’s not get ahead of ourselves… findings are mixed, with some studies showing no relationship between pre-pregnancy body mass index (BMI) and antenatal and postnatal depressive symptoms (e.g. Sundaram et al, 2012). So, can pre-pregnancy BMI affect risk of depressive symptoms during and after pregnancy? Dachew et al. (2021) attempted to answer this question in a recent systematic review and meta-analysis, which will be discussed in this blog.

This recent review and meta-analysis aimed to clarify the link between pre-pregnancy BMI and risk for antenatal and postnatal depressive symptoms.

Methods

The researchers conducted a systematic review following PRISMA and MOOSE guidelines. They searched four large databases for observational studies written in English reporting the association between pre-pregnancy BMI and maternal antenatal and/or postnatal depression and/or anxiety; search terms included “Pre-pregnancy BMI” and “depression”.

Studies were excluded from the meta-analysis if they:

Were non-longitudinal;
Did not use the World Health Organisation (WHO) BMI classifications of underweight, ‘normal-weight’, overweight, and obesity;
Provided insufficient data to calculate risk estimates and associated 95% Confidence Intervals (CI);
Provided combined data for overweight and obese women or for anxiety and depressive symptoms.

There were not enough studies on pre-pregnancy BMI and perinatal anxiety symptoms to conduct meta-analyses.

Two researchers independently assessed the eligibility, quality (using the Newcastle-Ottowa scale), and risk of bias (using the GRADE approach) of each study.

Data on the association between pre-pregnancy underweight, overweight, and obesity and antenatal and postnatal depressive symptoms were pooled from identified studies using odds ratios (OR) or risk ratios and 95% CI.

Results

The researchers identified 55 papers for full-text screening based on the initial screening of 6,788 records. A total of 35 eligible studies were included in the systematic review, and 13 eligible studies were included in the meta-analysis.

Antenatal depressive symptoms

Women with pre-pregnancy obesity had a 33% increased risk of developing antenatal depressive symptoms compared to ‘normal-weight’ pregnant women. Consistent findings were demonstrated when each study was individually removed from the analysis (leave-one-out sensitivity analysis).

There was no increased risk of antenatal depressive symptoms for women with pre-pregnancy underweight or overweight compared to ‘normal-weight’ pregnant women. Both pre-pregnancy underweight and overweight had considerable, meaning systematic differences between studies may have influenced results. The sensitivity analysis revealed a positive relationship between pre-pregnancy underweight and overweight and antenatal depressive symptoms when Laraia et al’s (2009) study was removed.

Meta-regression analysis looked at the proportion of variance in association between pre-pregnancy BMI and antenatal depressive symptoms explained by study location, sample size, outcome diagnostic measures, and study quality. These factors explained 66% of variance for underweight, 94% of variance for overweight, and 0% of variance for obesity.

Postnatal depressive symptoms

There appeared to be increased risk of postnatal depressive symptoms for women with pre-pregnancy underweight, overweight, and obesity compared to ‘normal-weight’ pregnant women. Leave-one-out sensitivity analysis revealed consistent findings, while between-study heterogeneity was not significant. The researchers claimed there were not enough studies to conduct subgroup analyses for the postnatal depressive symptoms.

Pre-pregnancy obesity increased risk of antenatal depressive symptoms and pre-pregnancy underweight, overweight, and obesity increased risk of postnatal depressive symptoms.

Pre-pregnancy underweight, overweight, and obesity increased risk of postnatal depressive symptoms.

Conclusions

So, what does all this mean? The researchers concluded that:

Women who were obese when they became pregnant had a 33% and 39% increased risk of antenatal and postnatal depressive symptoms, respectively, compared with ‘normal-weight’ women. Underweight and overweight women also showed higher levels of depressive symptoms during the postpartum period.

Women who were obese when they became pregnant had a 33% increased risk of antenatal depression, and a 39% increased risk of postnatal depression compared to normal-weight women.

Women who were obese when they became pregnant had a 33% increased risk of antenatal depression compared to ‘normal-weight’ women.

Strengths and limitations

This meta-analysis synthesised findings from multiple studies to provide a clearer picture of the relationship between pre-pregnancy BMI and antenatal and postnatal depressive symptoms. Nearly all studies in the meta-analysis were good quality (11/13) and had low risk of bias (12/13). Appropriate sensitivity analyses were conducted.

However, some limitations must be considered. The researchers only included papers written in English and the majority of studies were conducted in Europe and the US (10/13), so findings may not apply to other countries. Chen et al. (2018) found the highest rate of overweight and obesity during pregnancy in India, yet no studies from Indian populations were included.

As mentioned, when Laraia et al. (2009) was removed from the analysis, results changed. While investigating, I noticed Laraia and colleagues used the ICD’s BMI classifications, not the WHO’s (despite this being an inclusion criterion), with a higher cut-off for underweight, overweight, and obesity. So, findings may not have been comparable to other studies in the meta-analysis. This also makes me wonder how carefully exclusion criteria were checked (a vital part of any research process…), potentially contributing to between-study heterogeneity and limiting the meta-analysis’ reliability and validity.

It’s not ideal that the papers included in the meta-analysis controlled for different sociodemographic characteristics, with many failing to adjust for important confounders (e.g. only one adjusted for history of depression). The researchers could have listed factors that studies had to control for to be included, or taken these differences into account in analysis.

Publication bias was found for studies investigating obesity and postnatal depressive symptoms, but the researchers later claim they found no publication bias (interesting…). Duval and Tweedie’s (2000) Trim and Fill procedure should have been used to calculate an adjusted effect size for these studies, taking publication bias into account.

Findings may not be generalisable to women who live in countries outside the US and Europe, including low- and middle-income countries.

Findings may not be generalisable to women who live in countries outside the US and Europe, such as low- and middle-income countries.

Implications for practice

This meta-analysis highlights the need to ensure that mothers who are obese before pregnancy are screened for antenatal depressive symptoms throughout pregnancy, as they may be a high-risk group. Healthcare professionals should pay attention to nutrition among these women during pregnancy, as deficiencies, like low levels of omega-3 fatty acids, have been linked with increased antenatal depressive symptoms (Golding J. et al, 2009). Women with pre-pregnancy underweight, overweight, and obesity appear to be at increased risk of postnatal depressive symptoms. Safety permitting, healthcare professionals could encourage overweight or obese women to engage in physical activity while pregnant. As of April 2020, GPs must offer women a postnatal consultation 6-8 weeks after birth. This is not enough; ideally consultations should be offered regularly for at least a year, as postnatal depressive symptoms can begin at any point the year after birth (NHS, 2021). Women with pre-pregnancy underweight, overweight, or obesity should be particularly encouraged to attend these consultations due to their heightened risk for postnatal depressive symptoms.

The message is clear: mental health screening should become a key part of antenatal and postnatal check-ups, to ensure women receive necessary support and can provide optimal care to themselves and their new-born. Clinicians should be particularly vigilant when checking in with women at greater risk of experiencing perinatal depressive symptoms due to their weight.

It would be helpful to conduct a meta-analysis investigating the association between pre-pregnancy BMI and antenatal and postnatal depressive symptoms taking sociodemographic factors into account. This meta-analysis did not look at the effects of confounders controlled for (e.g. income, marital status, mental health history) on the relationship between pre-pregnancy BMI and antenatal or postnatal depressive symptoms. We must explore these potential risk factors to identify women who are at even greater risk of experiencing depressive symptoms during and after pregnancy. This could also help explain why pre-pregnancy BMI might be linked with perinatal depressive symptoms, exploring potential mediators and confounders impacting this relationship. Such research would benefit from including studies from around the world to ensure generalisability. The fact that there were not enough studies investigating the relationship between pre-pregnancy BMI and perinatal anxiety symptoms to conduct a meta-analysis emphasises the need for more research in this area.

There is work to be done when it comes to identifying the link between pre-pregnancy BMI and perinatal mental health problems. However, this meta-analysis takes us a step closer to understanding groups who appear to be at greater risk of developing debilitating conditions.