A Comparison of Breastfeeding Experiences between Mothers Spending the Traditional Chinese Confinement Period in a Confinement Centre or at Home

Background Ethnic Chinese mothers in Malaysia adhere 30 of traditional post-partum practices (the “confinement period”) aimed at recuperation after delivery. Recently there has been an emergence of confinement centres (CCs), where mothers stay and receive traditional confinement care. Ethnic Chinese mothers have low breastfeeding rates. There are concerns that practices in CCs could contribute to this but no data exists. We describe mother’s breastfeeding experience at CCs and identify areas for potential improvement in breastfeeding support. Methods Ethnic Chinese mothers were recruited telephone interviewed Chi-square test was used to compare groups and multiple logistic regression was used to assess the effect of confinement place on exclusive breastfeeding.


Conclusions
Contrary to current perceptions, breastfeeding experience and breastfeeding rates were similar at CCs and home but there were gaps in how CCs supported breastfeeding. Targeted training to CC staff to support breastfeeding may result in better outcomes for mothers staying in CCs.

Trial registration
National Medical Research Register NMRR-17-1174-36384S1 Background Breastfeeding is the cornerstone of infant nutrition and the dangers of not breastfeeding are well recognised (1). The early initiation of breastfeeding after birth and subsequent early post-partum establishment of breastfeeding are important for successful breastfeeding (2). Close and continuous proximity of mother and baby 24 hours a dayis also important to the establishment of breastfeeding (3).In some traditional cultural post-partum practices,restrictions to contact between mother and baby may occur. These may affect mother-infant interaction and breastfeedingpractice.
Malaysia has a multi-ethnic societyandeach ethnic group hasits own traditional postnatal practice.Chinese ethnicity is one of the 3 major groups and despite modernisation, most Chinese women adhere strictly to a 30-day traditional post-partum period, known locally as the "confinement period" or "zuoyuezi' (4)(5)(6)(7).
These practices are based on the traditional Chinese belief system about the maintenanceof Yin and Yang in the mother after childbirth so that her health is restored. Many believe that non-adherence to these practices will result in potentially long-term adverse effects on the quality of life to the mother.
Influences of relatives, in particular the new mother's mother, her mother-in-law and her grandmother, are factors that ensured that these practices are carried out and passed down from generation to generation (4,7).
During the confinement period, the mother would be assisted fully at her home by someone (or sometimes more than one person) -ensuring that she gets enough rest, providing the appropriate diet, caring for the newborn baby and ensuring she abides by confinement practices such askeeping warm(8). This usually means the motheravoids draughtsor taking showers for the whole month (4,7,9). This person was traditionally a close female relative (6)but practices have evolved to hiring a professional 'confinement lady' (yueso) who is considered an expert in the necessary diet and practices (9,10). The confinement lady would be employed to stay in the new mother's home after delivery for at least 4 weeks(10).
Over the last decade, a new method of providing this care emerged and is increasingly replacing the role of confinement ladies or female relatives at home. Confinement centres(CCs)provide a place for post-partum Chinese motherstostay during their confinement period and observe traditional postpartum practices under the care of theCCstaff (10). A typical CC is often a converted house with several rooms for mothers and a room designed as a nursery where the babies are all placed. Mothers may share rooms and typically all the babies would be in a single 'nursery'. This centre would be staffed by women who have experience with newborns and are familiar with the Chinese cultural confinement requirements, a cook to prepare the required confinement diet for mothers and occasionally qualified nurses to look after the well-being of mothers and babies (10).Although care of the infant is in no way neglected, but because the focus of these CCsis mainly towardsproviding traditional care for the mother, compromises in mother-infant bonding and breastfeeding may occur.
This has led to concerns that breastfeeding may not be adequately supported in CCsand therefore,further compromisingbreastfeeding rates for the Chinese population which is currently the lowest among the major ethnic groups in Malaysia (11,12) despite evidence that a high proportion of them intend to breastfeed (11).
The aim of the study is to describe the breastfeeding experience of Chinese mothers who stayed in CCs and compare the breastfeeding outcomes at 1 and 6 months with Chinese mothers who stayed at home to determine if the place of confinement had any effect on breastfeeding.
What this study will add: There is currently no published data about breastfeeding practices in confinement centres and these centres are a growing phenomenon across Asia and beyond. This study describes current practices at confinement centres and identifies areas for potential improvement in breastfeeding support in confinement centres.

Study Design
This was a prospective cohort study. Participants were recruited from the postnatal wards of six hospitals in Penang, Malaysiabetween August 2017 to October 2017.Written consentwas obtained from the participants prior to the commencement of study.

Participants
Participants were Malaysian of Chinese ethnicity, married and aged above 21 years, who delivered a healthy full-term infant within the previous 48 hours and had the intention to breastfeed. Recruitment was done by the infant's attending doctor, who apart from this was not otherwise involved in the study. For every recruited participant who chose to go to a confinement centre, a control (i.e. a mother who chose to stay at home) was recruited for the purpose of comparing breastfeeding outcomes.

Data collection
After consent was obtained, maternal and infant baseline characteristics were collected before discharge. These included maternal age, educational level, occupation, parity, previous breastfeeding experience, place of intended confinement; infant's date of birth, hospital of birth, gender, gestational age at birth, mode of delivery and birth weight. After discharge, there was no contact between the research team and the participant until after her30-day confinement period.
A telephone interviewwith the participant was conducted at 1-month post-partum (the end of the confinement period). The focus of this telephone interview was to seek the breastfeeding experiences of participants who stayed at CCs and compare it with those at home.The questions included infant's feeding practices, opportunity to room-in and/or spend time with their infants, problems encountered during breastfeeding (such as engorgement, mastitis, insufficient milk), sources of help for these problems and any perceived barriers to breastfeeding.There were also free field options for participants to make comments on their experiences. The second telephone interview to collect data on breastfeeding practices was conducted when the infant was 6 months old. All questions used had been tested in a separate group of breastfeeding mothers not involved in the study. Both telephone interviews were conducted by 3 trained research staff and the responses were directly entered into a specially designed interview form.

Sample size calculation
Based on the data on exclusive breastfeeding rates at 0 to 2 months for Malaysian Chinese mothers obtained from the National Health and Morbidity Survey 2016: Maternal and Child Health findings (13) and our hypothesis that there is a 50% reduction in breastfeeding rates in mothers who spend their confinement period in a confinement centre, we calculated the sample size using the Openepi software (14). The calculated sample size was 94 in each group,considering a 20% drop out rate.

Breastfeeding definitions
We defined exclusive breastfeeding according to the World Health Organisation's definition which means no other food or drink, not even water, except breast milk (including milk expressed or from a donor) for the first 6 months of life.Any breastfeeding was defined as breast milk (including milk expressed or milk from a donor) for 6 months of life, and any other food or liquid including non-human milk and formula (15).

Data analyses
We tabulated thebaseline demographics of the mothers according to place of confinement.
Continuous data was presented as mean with standard deviation (SD) and categorical data presented as frequency with percentage (%).Chi-square analysis was used to compare the baseline characteristics between participants staying in confinement centres (CCs) and those staying at home.Comments from free fill options were tabulated and categorized into groups. Some of these free field responses were quoted as illustrations.Simple logistic regression was used to determine whether CCs affected breastfeeding rates (any breastfeeding and exclusive breastfeeding) at 1 month and 6 months post-partumand presentedas crude odds ratio (OR) with 95% confidence interval (CI).Wemodeled the likelihood of exclusive breastfeeding or any breastfeeding as a function of CCs after adjusting for others clinically importantvariables. The results were presented as adjusted odds ratio (aOR) with95% CI. We considered a p-value of <0.05 as significant.

Results
A total of 187 mothers consented to participate, of which 88 (47%) chose to stay in a CC. Of the 99 (53%) participants who chose to go home during their confinement period, a third of them employed confinement ladies to provide them confinement care at home while remaining received their care from family members. We were unable to recruit a control for every mother who went to a CC because a greater proportion chose to go to their own homes.
We were able to interview 77 (87.5%)participants from the CC group and 96 (97%) from the home group at one-month post-partum. At 6 months, 68 (77.3%) from the CC group and 86 (86.7%) from the home group completed the interview. (See Figure 1)

Baseline characteristics
The maternal and infant baseline characteristics are presented in Table 1. The overall mean maternal age was 32.3(SD 3.97) years with 54% having previous experience with breastfeeding and 57% delivering in aBaby Friendly Hospital Initiative (BFHI)accredited hospital. The overall mean infant gestational age was 38.7 (SD 1) weeks and mean birth weight of 3149 (SD322)g. There were no differencesin the age, education background, delivery at a BFHI accredited hospital, mode of delivery; infant gestation and birth weight between the two groups. However, we found that significantly more primigravidas(53% CC vs 34% H, p < 0.01) but fewer mothers with prior breastfeeding experience went to CCs (44% CC vs 62% H, p<0.01). . We were unable to estimate the odd's ratio for 'any breastfeeding rates' at 1 month because all except 2 participants from the home group were breastfeeding (see Table 2). 1.0 (0.5 -1.9) 0.99 1.08 (0.5 -2.2) aAdjustedfor maternal education level, past breastfeeding experience, place of delivery, mode of delivery, spent more than 6 rooming in at night, employment status at 6 months b All mothers except 2 from the home group were practicing some form of breastfeeding Amongthe participants who werestill breastfeeding at 1 month (n=171), 29% were breastfeeding directly from the breast,49% onlyfed their infants with expressed breastmilk(baby was never latched to breast) and 21% used acombination of both. Significantly more participants from CCswere only feeding expressed breastmilkcompared to those from home (62% vs 39%, p <0.001) at 1 month (See Table 3). We explored reasons for not directly breastfeeding at 1 month. The main reasons were:participants perceiveda need for undisturbed rest and that direct breastfeeding was too time consuming (n = 28);had breastfeeding problems mainly inability to get baby to latch and sore nipples (n=22);had difficulty accessing their baby in the nursery or were discouraged to directly latch their infants(n=8).
Sevenparticipants from each group felt that they needed to monitor baby's milk intake. One mother cited that she did not want to place the baby near her breastswas because "I feel dirty because I haven't had a showerfor so long"

Breastfeeding experience
Overall, 66% of the participants in the CC groupreported that their centre was supportive of breastfeeding. When asked reasons for saying so, they cited one or more reasons which we categorised into the following: staff helped them overcome breastfeeding problems, mental support from staff, breastfeeding education given by staff (especially for mothers with no prior breastfeeding experience), availability of peer support from other mothers in the same centre. Of note, none of the participants stated that they perceived their centreto be supportive of breastfeeding because they had received encouragement to spend time with their infant or to breastfeed at night.
For both groups, we asked if they faced any hindrance to breastfeeding. Ten participants from the CC groupreported that they did encounter one or more forms of hindrance which we categorized into the following: staff were not helpful with breastfeeding problems; staff encouraged formula feeding; mother was asked to pump less breast-milk because there was inadequate refrigerator storage space; inappropriate informationgiven regarding breastfeeding such as the need to stop breastfeeding in a jaundiced baby or for diarrhea or skin rashes; or misconceptions such as "baby would be hungry without formula", "breastmilk will cause indigestion", "breastfeeding was troublesome and would result in a clingy baby"; family members against breastfeeding. However, not being allowed to breastfeed at night was not cited as a perceived hinderance to breastfeeding. Interestingly, three of the participants who said they had some hindrance to breastfeeding had earlier reported that they found their centresto be supportive of breastfeeding. On the other hand, 8 participants from the home group indicated that they had faced breastfeeding hindrance either because their family members or hired confinement lady did not support breastfeeding.
Sleeping arrangements during the confinement period Only 6(8%) participants from the CC group slept in the same room as their babies in contrast to 66(69%) from the home group (p = 0.00).Most of the participants in the CC group(n=71) slept separately from their babies: the mother slept in a mother's room (sometimes with a few other mothers in the same room) while their babies slept in a nursery. However, only 44% of participants from the CC group said that their centres did not allow them to sleep with their baby. The remaining said that they could but had chosen not to because they wanted to rest.
Nurseries in CCs were generally small and crowded with limited space to move around and there was usually nowhere where a mother could sit and breastfeed or spend time with her baby.In fact, it was a common practice that mothers would not be allowed to enter the nursery.However, for those that were allowed to enter,73% said they could access their babies at anytime of the day or night, whilethe remaining 27% had some restrictions. In particular, 17 participants reported that the centre discouraged or disallowed breastfeeding at night. Other restrictions included designated timesto be with or seetheir baby through a window in the nursery. SeeTable 4 for more details of the type of access available. Table 4 Degrees of access to babies (for participants who did not sleep with their babies, n=71) Description n (% Baby could be brought out of the nursery at any time to be with mother. In addition, mothers could see their babiesfrom outside of the nursery through a window, any time of the day or night. 51 (6 Baby could be brought out of the nursery at any time to be with mother. Window for mothers to see their babies in the nursery only open for certain hours of the day. 8 (10 Baby could be brought out of the nursery at any time to be with mother butunable to see baby while in the nursery. 2 (3) Baby allowed out of the nursery during the day. No physical access to baby at night. However, mothers can see their babies through a window from outside of the nursery any time of the day or night.

(6)
Baby only allowed out of the nursery during the day. No physical access to baby at night. Window for mothers to see their babies in the nursery only open certain hours of the day.

(4)
Baby only allowed out of the nursery during the day. No physical access to baby at night. Unable to see baby from outside of the nursery.

(3)
Breastfeeding problems encountered and sources of help Breastfeeding problems were experienced by 81% in the CC group and 76% in the home group.
Among those with breastfeeding problems, significantly more participants in CCs (79%) compared with those at home (59%) encountered latching difficulties, breast engorgement, blocked ducts, mastitis, abscess and sore nipples(p = 0.03). However, there was no difference in the number of participants experiencing inadequate milk (n=40 vs 33, p = 0.06).The number of participants who perceived that family members were the main source of hindrance to breastfeeding was also similar in both groups (n = 9 vs 8, p = 0.64).
Sources of help for breastfeeding problems were different for participants in CCs and home.
Participants in CCsmainly sought help from centre staff (39%) and breastfeeding support groups (19%); while most home participantsobtained help either from friends (17%)or healthcare professionals (20%). Only 7% of participantsstaying at home obtained help from family members and of the 32 participants who hired a 'confinement lady', 8 sought their assistance. For both groups, only 4%sought help from lactation consultants.

Discussion
We found that breastfeeding experience were similar and exclusive breastfeeding rates at 1 month and 6 months were not different between mothers who went to CCs and those who stayed at home.
This is contrary to current perceptions that CCsare a cause of early breastfeeding cessation.However, it should be noted that all mothers included in this study had the intention to breastfeed and having such anintention is a recognized factor affecting breastfeeding duration (16). We believe that thisintention was the main factor why mothers in this study persevered with breastfeeding despite presence of some practices in CCs that hindered breastfeeding.Another reason for this finding was the perceived support that the mothers had from their CCs, even though there were gaps in the provision of support and a lack of understanding ofwhat appropriatebreastfeeding support actuallyentails. For example, despite limitations in access to their babies or being discouraged from rooming-in or feeding at night (all of which are knownbarriers to breastfeeding), these mothers felt that their CC was supportive of breastfeeding.There is evidence that mothers who perceive that they are supported are more likely to successfully breastfeed(3).
A major difference between the CC and home group was that there was significant separation of mother and infant. Only 8% of CC mothers slept in the same room with their babies compared to 69% of mothers who stayed in their own homes. The practice of separating mothers from their babies at CCs was most likely due to the cultural emphasis on the need for postnatal mothers to rest during the confinement period (4,7,9).We found some CCs that discouraged mothers from rooming-in with their babies at night and this is worrying because the baby's early feeding cues could easily be missed, leading to a reduction inbreastfeeding frequency and hence the amount of breast milk a mother produces.Sleeping in the same room as the baby or 24-hour rooming-in is crucial for night feeds, which is important for the establishment and continuation of breastfeeding (3) We also found that many mothers had restricted opportunities to access their babies both during the day and night, and this could explain why significantly more mothers in CCsonly fed expressed breast milk. With a growing body of evidence that feeding expressed breast milk might be different from feeding at the breast, this is an important finding. Studies have found that there is higher risk of asthma (17)and childhood obesity (18)(19)(20)with expressed breast milk feeding compared to direct breastfeeding. There are also concerns that only feeding expressed breastmilk could result in shorter duration of breastfeeding (20).In our study,expressed breastfeeding in the first month of life did not make any difference to our 6-month breastfeeding rates, but our study was not designed to answer this question.
Apart from the above, the breastfeeding experience in the two settings wasotherwise similar. Mothers from both groups encountered people who hindered breastfeeding; some mothers found CC staff discouraged breastfeeding while those who stayed at home found their hired 'confinement lady' discouraged breastfeeding. In both the groups, family members were also cited as a barrier to breastfeeding.
We found that there were more first-time mothers in the CC group. We cannot directly explain this but possible explanations might include thelack of confidence to look after the baby themselves, orit might also be due to the increasing availability of CCs.
Our study is important because over the past decade there has been and continues to be a rapid expansion of CC services not only in Penang and throughout Malaysia but also in other parts of the world including Singapore, Hong Kongand mainland China (21,22). Similar centres are also opening to cater for the traditional needs of the other Malaysian ethnic groups.Therefore, this study offers insight into the level of breastfeeding support that is currently being offered and the findings will be useful in determining how support can be offered to these centres and to mothers at home. We noted that only a few mothers sought help from lactation consultants or peer support groups. There is currently no formal recognition of lactation consultants in Malaysia and the number of certified consultants is very low. Peer support groups exist but again our data showed that these were not sought either by the CC or home participants. Efforts to link CCs and mothers at home with these groups are needed. This study also provides insight into the training needs of the CC staff.
A limitation of our study would be that our findings reflect only the views of mothers who at recruitment had the intention to breastfeed. These findings do not apply to mothers without an intention to breastfeed, and such studies are needed. In addition, this study is limited because we did not have data on which CC the mothers went to. However, while we believe that our sample of mothers reasonably represents the mothers using CCs in Penang, we do not believe that we have a reasonable representation of the CCs in Penang. It is likely not all CCs in Penang were represented in the data and it is possible that several of our participants could have been at any one of the CCs.
Therefore, our results might not be generalizable to all mothers attending all CCs in Penang.

Conclusion
We found that spending the traditional confinement period in CCs or at home did not result in a difference in breastfeeding rates at one and 6 monthspost-partum among Chinese mothers who intended to breastfeed. However, when describing their experience, we found definite gaps in how CCs were supporting breastfeeding.
Ways to overcome this need to be explored. One such intervention might be to provide training to improve CC staff's ability to provide breastfeeding support. Such training would need to be tailored specifically to their needs and thus CC staff input into the design of an educational training package is recommended. Future studies could also be done on all mothers regardless of their intention to breastfeed.
List Of Abbreviations