The main findings in the present study were that ward routines/practices at both the labour and the maternity ward influenced milk ingestion/production four days after birth. In addition, maternal experience of breast engorgement and feeling "low/blue" were related to milk ingestion/production. The influence of these variables differed between primi- and multiparous women.
In the current study, milk production/ingestion four days after birth or the duration of "nearly exclusive breastfeeding" did not differ between the groups allowed skin-to-skin contact, being in mother's arms dressed or swaddled, or exposed to a short-term separation after birth. In contrast, the procedure of early suckling was associated with an increased milk ingestion/production four days after birth.
In some, but not all, previous studies [7, 18], positive effects of early skin-to-skin contact on the duration of breastfeeding have been demonstrated [19–22]. Early suckling has also been shown to promote breastfeeding [19, 23–25]. The different results obtained between studies may be related to the fact that the design of the previous studies did not always allow a clear distinction between the effects of skin-to-skin contact and early suckling on the outcome of breastfeeding.
It should be mentioned, however, that the infants in our study that suckled early weighed significantly more and had a significantly higher mean gestational age than those that did not suckle, differences which may be related to the outcome.
Surprisingly the amount of milk ingested/produced four days after birth did not differ between the groups of mothers and infants that were allowed to stay together after birth (Skin-to-skin and Mother's arms groups) or that were subjected to a short-term separation (Reunion group), in spite of the positive effect of early suckling in the former groups.
Maternity ward routines had a strong impact on the outcome of milk ingestion/production in the present study. Infants staying in the nursery at the maternity ward ingested less milk than infants who were rooming-in and thus allowed breastfeeding on demand. This difference in milk ingestion reached significance in infants of multiparous women. This lower milk intake was not related to more supplementation.
The difference in the amount of milk ingested/produced on day four may be related to the fact that the number of breastfeeds was set to seven times per day in the group of mothers who had their infants in the nursery (Group III), to the use of infant formula in the nursery, and most likely to the more limited contact between mother and infant subjected to this routine due to the separation of mother and infant in between feeds.
Swaddling did not have any significant effect on milk production as had been hypothesised, since in a previous study we found that swaddling could delay recovery of weight loss in babies while combined with potential stressors, such as separation after birth and supplementation by formula or glucose .
The relative importance of the different parameters shown to be of importance for milk ingestion/production four days after birth was further investigated by means of regression analysis. The design of the study allowed four different analyses to be performed. Thus, groups of mothers and infants that were staying together at the labour ward and those that were separated at this time were analysed separately and there was a further split of the data based on parity (Table 4 and Table 5).
In the groups of mothers and infants that stayed together in the labour ward, early suckling, the intensity of breast engorgement and the number of breastfeeds on day 3 after delivery were positively associated, and the intensity of feeling "low/blue" was negatively associated, with the amount of milk ingested/produced four days after birth. These relationships were, however, only seen in the primiparous mothers. In multiparous mothers, early suckling was found to be the only factor of importance.
A similar pattern was obtained in the mothers and infants that were exposed to a short-term separation after birth. Thus, even in the separated groups, the intensity of breast engorgement was positively associated and the intensity of feeling "low/blue" was negatively associated with the amount of milk ingested/produced four days after birth in the primiparous women, whereas the routine of rooming-in was of singular importance for milk ingestion/production in multiparous women.
One conclusion that could be drawn from the regression analyses is that milk production might be differently influenced in primi- and multiparous women. Obviously, the intensity of breast engorgement was associated with milk ingestion/production in primiparous mothers, irrespective of labour ward routines, whereas multiparous mothers seemed to rely more on "sensory stimulation", such as the procedure of early suckling or the routine of rooming-in.
Breast engorgement has been suggested to be related to "hormonal" stimulation of milk production. Oestrogen and progesterone levels are high during pregnancy, but since these hormones are, to a great extent, produced in the placenta, the levels of these hormones decrease rapidly after birth. Decreased levels of these steroid hormones have been associated with initiation of milk production [26, 27]. Possibly the experience of physiological engorgement is related to this phenomenon.
Oxytocin is of main importance for milk ejection and related behaviours . Animal studies indicate that there are differences in primi- and multiparas regarding sensitivity to sensory cues that trigger milk ejection and maternal behaviour. Animal studies show that in ewes, the first pregnancy induces permanent changes with regard to the release of oxytocin and oxytocin receptors, e.g. the expression of oxytocin receptor mRNA is increased in the paraventricular nucleus (PVN) of the brain and the sensitivity of oxytocin autoreceptors is increased . These changes resulting from maternal experience may facilitate the expression of oxytocin-mediated functions, such as milk ejection, in connection with later pregnancies. An analogous finding is that the prolactin surge to suckling is blunted in rats having previous experience of lactation . Studies in humans actually indicate a similar increased sensitivity of prolactin receptors in multiparous when compared to primiparous women .
These data may indicate that the "machinery" for milk production might be primed by previous lactations also in women. As a consequence, it may be easier to trigger milk production by sensory stimuli. Thus, early suckling and/or rooming-in may suffice to initiate milk production, when the mammary tissue is already primed.
The results of the multiple regressions must, however, be interpreted with some caution. Since they were based on information obtained on days 3 and 4 of lactation, it cannot be excluded that other relationships might have been established if data from other time points had been explored. Other unknown factors may also play an important role. Still the similarity of the results obtained for primiparous women in the regressions performed on the groups of mothers and infants that stayed together after birth and those that were separated after birth speak strongly in favour of the reliability of the analyses.
In the present study we were able to cover the impact of some factors affecting milk ingestion/production four days after birth. This may not complete all possible influences such as cultural differences, for example. Yet the above data suggest that milk production, a physiological function of vital importance for the offspring and survival of the species, is under multifactorial control. When infants are allowed close contact with their mother immediately after birth, early suckling may be a strong promoter of milk production. If, on the other hand, mother and infant are separated after birth, e.g. because of complicated delivery, alternative physiologic pathways may be activated to ascertain an adequate production of milk. Intense sensory stimulation as offered by the routine of rooming-in or feeding on demand may compensate for effects cause by short-term early separations.
It is important to mention that early skin-to-skin contact should be promoted even if it does not seem to play a direct role for milk production and duration of breastfeeding. Skin-to-skin contact in this study started between 20 – 25 minutes after birth as the infants were subjected earlier to compulsory hospital practices. This may have interfered with the baby's developing breastfeeding behaviour  and may have disguised some effects of immediate skin-to-skin contact.
Even if there are no measurable effects of skin-to-skin contact, it may still create a positive experience for the infant and the mother and facilitate the procedure of breastfeeding [1, 32]. Additionally, stress levels have been shown to be reduced both in mothers and infants [3, 4], and, as has been demonstrated by Klaus and Kennell  amongst others, early skin-to-skin contact occurring immediately after birth may promote attachment between mother and child and facilitate maternal-infant interaction in a long-term perspective.