4.1. Discussion of results on population characteristics
Most of the women in our cohort were first-time mothers, consistent with existing literature [12, 16].
Approximately half of the women had a history of psychiatric disorders, also in line with previous research [12]. It is well-established that women with a history of mood disorders are known to have a significantly higher risk of developing postpartum depression (PPD) [17], which may explain the high proportion of women with a psychiatric history in our cohort. Six mothers had a history of substance use disorders and received follow-up care for addiction. It is hypothesized that women with prior psychiatric care are more likely to be referred to mother-baby units in the event of postpartum depression.
Two women were diagnosed with bipolar disorder in the postpartum. According to Tebeka, 10% of women develop bipolar disorder through post-partum depression [18].
The study found that a majority of the mothers had a comorbidity related to anxiety, which is consistent with previous research [10, 12]. Our study also found a similar prevalence of anxiety among mothers with postpartum depression as a French national study conducted in 2023, which reported a prevalence of 83.2% [19].
4.2. Results discussion
4.2.1. Primary outcome
The study's initial finding is a statistically significant decrease in both global DQ and partial DQ for sociability and posture in our population compared to the general population, without reaching pathological values.
These values indicate a trend towards a very early drop in DQ, necessitating special attention, monitoring, and support.
The statistically significant reduction in overall DQ in our cohort is consistent with the literature [20].
Therefore, we hypothesize that the lack of delay in the 6-month-old infants in our study may be attributed to the mother-baby care provided, which safeguards the child's development.
Infants of mothers with postpartum depression tend to avoid eye contact by frequently turning their heads, displaying reduced expression of positive affect and reduced psychomotor activity, which is used to distance themselves from interaction failures [21].
This relative weakness in sociability found in our cohort is in line with what is described in the literature [22–24].
Our population exhibited significantly lower postural partial DQ, even though we excluded infants with pathologies that could interfere with postural tone development.
Ajuriaguerra has described the tonic-emotional dialogue, which is the body-to-body exchange between the infant and caregiver. The pre-verbal child expresses their emotions through tonic dialogue, from the tension-relaxation angle. Maternal depression can affect a mother's motor skills, muscle tone, and even the way she carries her child. This, in turn, can impact the child's tonic responses [25], which may explain the moderate drop in posture DQ scores in this population.
Other studies have also demonstrated the impact of depression on infants' psychomotor development [16, 26, 27]. Postural development (which can be likened to gross motor skills) develops earlier and faster than oculomotor coordination (which can be likened to fine motor skills). This evolution over time could explain the disparate results of studies evaluating the impact of PPD on motor development at different ages.
Our study found no significant change in the partial DQ 'language' of 6-month-old infants whose mothers suffer from PPD. Previous research has shown that mothers with mood disorders tend to use less speech [28] and exhibit reduced voice modulation when addressing their child [29–31]. In their study, Kaplan [32] demonstrated a significant negative correlation between the mother's sadness and the expressive communication subscores of the Bayley test. Other studies have also found that the duration of postpartum depression symptoms is associated with a negative impact on the child's language [33–35]. The authors who discovered impaired communication in the infants of depressed mothers had conducted their study when the children were at least one year old. Therefore, it is possible that our assessment at 6 months of age was too early to detect a significant change in the partial DQ partial language. This prompts us to continue assessments at 12 and 18 months. Furthermore, as previously stated, care in the mother-baby unit may have mitigated the impact of maternal postpartum depression on the child's communication.
4.2.2. Secondary outcomes
Maternal depression occurring within the context of bipolar disorder has been found to reduce all of the children's development quotients, compared to children of depressed mothers without bipolarity. However, the partial postural development quotient was the only one that was significantly lower. This trend can be attributed to the more severe depressive symptoms experienced by women with bipolar disorder, which have a greater impact on the mother-baby relationship [36].
In cases of comorbid anxiety in depressed mothers, the children's global quotient and partial quotients - with the exception of language DQ - were significantly reduced. This was more marked for the sociability DQ. The mother's anxiety has a further impact on the mother-child relationship. Anxiety makes mothers even less available to interact in an appropriate manner with their child [37].
Babies with anxious mothers are reported to have impaired emotional regulation, which explains the drop in partial socialization quotient [38, 39], as babies at this age regulate themselves in the relationship with their mothers, but also the drop in partial postural quotient due to tonic-emotional dialogue [25].
Partial language quotient was improved in children whose mothers had comorbid anxiety. In cases of anxiety comorbid with depression, the mother is often less psychically slowed down than in cases of depression without anxiety, and we can hypothesize that she talks more to the child, she overstimulates the child who is placed in a perpetual “language bath”. This could explain the preservation of the partial language DQ.
4.3. Discussion of the Brunet Lézine scale choice
Assessing early childhood development can be done using several available scales.
The Bayley Scales of Infant Development - 3rd Edition (BSID-III) is the most widely used scale in the literature for assessing the development of preschool children (between 1 and 42 months) [40]. It consists of five scales: cognitive, language, motor, social-emotional, and behavioral. The cognitive, language, and motor scales are evaluated by the examiner, while the social-emotional and behavioral scales are evaluated by the parents through a questionnaire [15].
This tool was not calibrated on a French population nor available in French until 2022. As a result, it was not possible to use it to collect data on the babies in our study, which took place between 2012 and 2023.
To assess early childhood development, we naturally used the Brunet Lézine test (BL), described in the methods section, which is the second most widely used scale. This tool has good psychometric properties and can be used to assess the development of babies up to two years of age [13].
A correlation study was conducted between the Brunet-Lézine scale and the BSID-III, revealing moderate correlations between the gross motor score of the BSID-III and the posture quotient of the Brunet-Lézine scale, between the fine motor score of the BSI-III and the hand-eye coordination quotient of BL, and between the socio-emotional score of the BSID-III and the sociability quotient of BL. It is worth noting that the study was conducted on a population of 36 pre-term infants with a corrected age of 6 months. However, the BSID-III language subscore and BL language quotient did not correlate well [41].
4.4. Bias, limits, and strengths of the study
Possible ranking biases should be considered. It has been shown that providing psychological support for mothers can lead to significant improvements in the mother-child relationship [42]. In our sample, depressed mothers received psychiatric care before the children's development was assessed. In our unit, infants were supervised by childcare workers who provided appropriate stimulation and attentiveness to the child's needs, which can help offset maternal withdrawal. It is therefore possible that the treatment provided helped to reduce the developmental impact of maternal depression.
The study had some limitations.
Firstly, the number of infants studied was relatively small, which reduced statistical power.
Secondly, the study was conducted at a single center. However, the study was conducted in a referral center located in the second largest city in France, which serves a vast region in the south-east of the country. Furthermore, the mother-baby relationship care provided in the Marseille unit was consistent with that of other units in France [43].
Another limitation of the study was the inability to stratify women by socio-economic level due to a lack of information. Mothers with a higher socio-economic status tend to provide better care for their babies when suffering from PPD, according to Stein [6].
Our study's strength lies in the objective assessment of the babies by a healthcare professional external to the dyad's care, using a validated and parameterized scale. This reduces the measurement bias associated with the evaluator's subjectivity.