Difficulty bonding and postnatal depression are not the same thing
In 64,938 Japanese mothers, 7.7% reported bonding difficulty at one month after every woman with lasting postnatal depression had been excluded.
The last topic was about the first hour. This one is about the weeks after, and about a mistake that is made constantly in both directions: treating difficulty bonding as if it were simply depression, or treating it as if it never had anything to do with depression.
The study designed to answer exactly this. A Japanese national birth cohort registered 64,938 mother and child pairs between 2011 and 2014. Bonding was measured with a standard eight-question scale. Then the researchers did the thing that makes the study valuable: they removed every woman who had persistent postnatal depression, defined as scoring above the cut-off on the standard depression questionnaire at both one month and six months after birth. Among the women who remained, that is women without lasting postnatal depression:
Roughly 1 in 13 mothers, with every woman who had persistent postnatal depression taken out. Bonding difficulty without depression is real, it is common, and it has its own causes.
What was associated with it, with the baseline attached. Against that 7.7% starting point, the strongest single factor was finding it hard to hold the baby when the baby was cranky or arching its back (adjusted odds ratio 3.45, confidence interval 2.96 to 4.03). Negative feelings about the pregnancy itself came next (2.42, 2.20 to 2.65). The strongest protective factor was high social support (0.45, 0.41 to 0.49). Notice what is on that list: how the baby behaves, and how much help you have. Not character.
A second study measured the two things separately on the same women. In a French maternity ward, 78 mothers completed the bonding scale 48 hours after delivery and were also assessed by a child psychiatrist. The bonding scale performed well against that assessment (area under the curve 0.93). The finding that matters here: the bonding score was statistically independent of the depression score (r = 0.11, p = 0.29). It was, however, related to the baby's own behaviour (r = 0.3, p = 0.01). At 48 hours, how a mother felt about her baby was not a reflection of her mood score.
And now the other half of the honest answer. A clinical series of 44 women, all of whom had had at least one episode of postnatal mental illness, described severe failure to love a particular baby. Its conclusion, quoted exactly: Postnatal mental illness and recalled severe pain during labour were significantly associated with such disorders which, in their severe forms, did not occur in the absence of postnatal mental illness.
That series carries no prevalence figure and none is taken from it. What it establishes is a boundary: at the severe end, in that series, bonding failure did not occur without postnatal mental illness.
Put it together and the accurate sentence is neither of the two you usually hear. Difficulty bonding is not a symptom of depression, and it is not unrelated to it. In the ordinary form it occurs commonly without depression and is separately caused. At the severe end, in the one series that examined it, it was always accompanied by postnatal mental illness. Both of those are true at once, and that is precisely why the two things should be asked about separately rather than one being used as a proxy for the other.
Two practical notes. The bonding questionnaire used in these studies is short, eight items, and was designed to be filled in on day 3; scores at 3 days and 12 weeks moved together in the original work. And in the French study, nurses' impressions of how mothers felt agreed only weakly with what the mothers themselves reported (r = 0.31), with poor agreement item by item. The authors concluded: New mothers need to express their feelings about their babies, as hospital staff observation of mother-infant interactions is not sufficiently reliable for assessing the attachment process.
Nobody can tell how you feel by watching you. You have to be asked, and you have to be able to answer honestly.
The limits. The large modern figures here are Japanese and the validation study is French. No large modern US cohort measuring this could be found for this guide, and none is invented.
"I am finding it hard to feel connected to him. My mood is fine. Is that something you see?"
"I filled in a mood questionnaire. Did anyone score it, and what did it say?"
"Can we set a time to talk about this again in a few weeks?"