The Effect of Intermittent Kangaroo Mother Care on Breastfeeding Practice of Late Preterm Infants in China

China has an extremely low exclusive breastfeeding rate. Kangaroo mother care (KMC) has been shown to increase the exclusive breastfeeding rate of babies born extremely or very preterm. However, there is limited evidence of intermittent KMC’s effect on exclusive breastfeeding in late preterm. Using hospital discharge and follow-up data collected from four postnatal wards, we investigated the association between intermittent KMC and breastfeeding practice for late preterm babies in China. for

Mothers in the KMC group were more likely to be breastfeeding (method) than mothers in the no-KMC group (65.3% vs. 52.1% at discharge, and 83.1% vs. 67.3% at follow up). Logistic regression indicated that compared with the no-KMC group, mothers who provided KMC were twice as likely to be providing their newborns with exclusive breastmilk feeding at discharge (OR=2. 15 (1.53-3.02)), use breast feeding method at discharge (OR=1.61 (1.15-2.25)), provide exclusive breastmilk feeding at follow-up (OR=2.55 (1.81-3.61)), and use breast feeding method at follow-up (OR=2.09 (1.44-3.02)).

Conclusions
Intermittent KMC was associated with a nearly doubled increase in exclusive breastmilk feeding (outcome) and breast feeding (method) at both discharge and 42 days after discharge for late preterm babies. Intermittent KMC has the potential to increase exclusive breastfeeding rates amongst late preterm infants.

Background
Kangaroo mother care (KMC) is a cost-effective intervention recommended by WHO for the care of stable preterm newborns (1). Infants are cared for skin-to-skin on the chest of their mother or another caregiver and receive exclusive breastfeeding (ideally) or breastmilk feeding. Compared to conventional care it has been shown to reduce mortality, the incidence of severe infection and hypothermia and improve health outcomes including exclusive breastfeeding. KMC has been found to increase the likelihood of exclusive breastfeeding at hospital discharge by 50% and at 1 to 4 month follow-up by 39% (2).
The breastfeeding rate in China is low, for both term and preterm newborns.
According to a survey of over 10,000 mothers in 2018, the rate of exclusive breastfeeding for infants under 6 months of age was 29.2%, and only 11.3% of mothers breastfed their children within an hour of birth (3). While the survey did not disaggregate breastfeeding rate based on gestational age or birth weight, a more recent observational study with a smaller sample size indicated an exclusive breastfeeding rate of 22.5% at 6 months in infants born preterm (4), lower than their full-term counterparts. Barriers to exclusive breastfeeding of preterm infants in China include low levels of breastfeeding self-efficacy and symptoms of postpartum depression amongst mothers. Additionally limited knowledge among health care providers regarding breastfeeding preterm infants and the specific benefits of breastmilk for these newborns hindered their ability to support new mothers (4,5). There is empirical evidence of the benefit of breast milk for preterm infants including improved cardiovascular function in adulthood (6).
While there is evidence to demonstrate that KMC increases the exclusive breastfeeding rate amongst preterm infants, most of this evidence was generated from outside the Western Pacific region (2,7,8) and there have been no studies conducted in China. The majority of these studies are set in neonatal units with extremely preterm or very preterm newborns. The effect of intermittent KMC on the breastfeeding outcome of late preterm newborns cared for on postnatal wards, characterized by rooming in of mother and baby and limited length of hospital stay needs investigating.
In this study, we aim to determine the effect of intermittent KMC exposure for late preterm newborns in a postnatal ward setting on exclusive breastfeeding at hospital discharge and follow-up. This analysis is part of a larger project to investigate the implementation of KMC in China as part of a national-level pilot.

Study design and population
This study is part of a larger piece of implementation research on the practice of KMC was recommended to the mothers and families of preterm babies on the postnatal wards. Exclusive breastfeeding was also recommended regardless as to whether mothers chose to provide KMC. The study was explained to mothers and families and written consent for participation was sought. In total 1,007 mother and baby pairs were enrolled in the study. Preterm twins were recorded as two separate pairs in the analysis. Those who chose to provide KMC were termed the "KMC group" (N = 752) and those who chose not to provide KMC were called the "no-KMC group" (n = 255). The specific KMC provision flow diagram is shown in Fig. 1.
Basic maternal socio-demographic information was collected using a structured survey along with obstetric history including maternal parity, presence of obstetric complications and mode of delivery. The newborns birthweight and gestational age were recorded. During hospital stay feeding outcome and method were recorded daily, and each episode of KMC provision was recorded and its length noted. At hospital discharge feeding practice (outcome and method) for the preceding 24 hours was recorded. A follow-up survey was conducted by phone for all participants 42 days after hospital discharge. The survey included questions about feeding practice (outcome and method) provision of KMC after discharge, the incidence of serious disease in mother or newborn and the current weight of the infant.
Ethical approval for the study was obtained from Peking University First Hospital Biomedical Research Ethics Committee. All participants gave written consent.

Measures and variables
The main outcome variables analyzed were the feeding outcome and the feeding method of preterm infants 24 hours before discharge and 42 days after discharge.
Feeding outcome refer to the make-up of the infant's feeds, classified as "exclusive breastmilk feeding", "exclusive formula feeding" or "mixed feeding" (when a baby receives both breast and formula milk), while feeding method refers to the way in which newborns receive their milk, this was classified as either "breastfeeding" (breastfeeding, or breastfeeding and other fed) or "other" which only included babies fed via bottle, tube, syringe or cup and no breastfeeding.
The independent variable in the analysis was KMC vs. no-KMC. We also included other socio-demographic indicators as exposure variables in the study, including maternal age, education attainment (high school, college, university and above), occupation (government employee, technician, worker, etc.), parity (primipara or multipara), pregnancy-related complications (yes, or no), infant's birth weight ( > = 2500 g, or < 2500 g) and gestational week (36 weeks, or less than 36 weeks).
Additionally, for the KMC groups, we documented the total KMC session numbers and duration of each KMC session during hospital stay.

Statistical analysis
Selected socio-demographic and delivery-related variables were considered separately for the KMC and no-KMC groups. The average KMC frequency and KMC duration were distinguished for the KMC group only. To compare the difference in feeding outcome and feeding method between KMC and no-KMC groups, the unadjusted percentage of breastfeeding outcome and method at discharge and 42 days after discharge was calculated.
Logistics regression was used to analyze the association between KMC and feeding outcome and method. All models were commonly adjusted for covariates including age, education, occupation, parity, pregnancy-related complications, mode of delivery, birthweight and gestational age. Odds ratio for feeding outcome at discharge and at follow-up, and feeding method, were reported within KMC and no-KMC groups. All statistical analyses were performed using Stata V14, and SAS V.9.3, and test results were reported to be significant at 0.05 level. Table 1  Similarly, there were more mothers that delivered through C-section in the no-KMC group (69.6%) than in the KMC group (60.5%). 12.8% of mothers in the KMC group delivered before completing 36 weeks gestation, significantly higher than the no-KMC group (3.2%). KMC was provided an average of 3.5 times before hospital discharge and the average duration each time was 65.8 minutes.

Discussion
This study is the largest and one of only a few studies on KMC and breastfeeding in China. Our analysis shows that KMC was associated with a nearly two-fold increase in exclusive breastmilk feeding (outcome) and breast feeding (method) at both discharge and follow-up in late preterm infants. Our results suggest that for late preterm newborns on postnatal wards (with mother), relatively brief exposure to intermittent KMC in facility was associated with increased exclusive breastfeeding at discharge and at follow up.
Preterm infants are at a higher risk for late breastfeeding onset and early breastfeeding cessation when compared to infants born at term (9), due to a series of barriers including but not limited to a lack of adequate breast milk, an immature uncoordinated sucking pattern and increased likelihood of maternal symptoms of depression due to preterm delivery (4,10,11). This also applies to late preterm infants (12,13). Inadequate milk intake contributes to slow weight gain and protracted jaundice in late preterm infants, making routine formula supplementation and early termination of breastfeeding more likely (14).
Our study is consistent with previous evidence from published literature that frequent skin-to-skin contact between mother and baby is crucial to the successful transition to direct breastfeeding in preterm infants (7,11) and initiation of exclusive breastfeeding in healthy full-term babies (15). Early skin-to-skin contact, within the first hour of birth, if possible, facilitates maternal milk production (16,17). While continued skin-to-skin contact on a daily basis accelerate neurophysiological development of the preterm infant (18), which contributes to establishment of effective suckling behavior. KMC on postnatal wards minimizes mother-infant separation time and likely increase breastfeeding (4,19).
It is noteworthy that KMC may play a role in alleviating stress related to preterm birth, encouraging mothers to care for their late preterm infants and breastfeed.
Parents of late preterm infants are likely to exhibit a lack of confidence and some may exhibit distress during feeding or symptoms of depression (20)(21)(22) Despite being the largest study on KMC and breastfeeding in China, there are several limitations to be considered. The study was not a randomized controlled trial, as KMC is known to be beneficial for preterm newborns it was deemed unethical to randomize mothers and babies to a group where they would not be encouraged to practice KMC. It is possible that those who opted to provide KMC may have been more likely to breastfeed their babies than those who chose not to provide KMC; however, we found no significant difference in socio-demographic characteristics (including age and educational attainment) between the two groups.
Moreover, the exclusive breastfeeding rate for mothers who did not provide KMC (33.2% at 42 days follow-up) is similar to the national exclusive breastfeeding rate at six months (29.2%); therefore, we believed that the finding that KMC was associated with an increase in breastfeeding is valid.
Secondly, we also noted difference in the KMC vs. no-KMC ratio across four study hospitals especially where in one hospital almost all of the participating mothers chose to provide KMC to their babies. We also did an additional analysis excluding this hospital and the results were similar (see supplementary appendix).
Nonetheless, we believe that the four hospitals enrolled are all tertiary hospitals with minimal difference in service delivery capacity, and it would be appropriate and useful to compare breastfeeding pattern between mothers in different hospitals, thus we included all of them in our analysis.
A third limitation is that, those who were not successfully followed-up were excluded from the analysis, which may lead to selection bias. However, the loss to follow-up rate was similar between in the KMC and the no-KMC group (15.4% vs. 14.5%). Within the KMC group, those lost to follow-up reported a higher exclusive breastfeeding rate at discharge than those who were successfully followed-up (60.0% vs. 54.6%), while within the no-KMC group those who were lost reported a lower rate than those who were followed-up (23.5% vs. 34.6%). This suggests that the results of our analysis may underestimate the association between intermittent KMC and improved breastfeeding outcomes as the actual impact may be greater.
Lastly, the major outcome variables of breastfeeding were only collected at hospital discharge and at 42 days follow-up. The study would have benefited from a longerterm outcome variable, e.g. exclusive breastfeeding at 6-months of age, in order to provide a more robust clinical and public health recommendation.

Conclusion
In this observational study, we found that KMC was associated with a nearly two-fold increase in exclusive breastmilk feeding (outcome) and breast feeding (method) at both discharge and at 42 days follow-up in late preterm infants. We believe this demonstrates the benefit of even "low-dose" intermittent Kangaroo mother care in late preterm infants. Additionally, considering the extremely low exclusive Figure 1 Operational process for Kangaroo mother care on postnatal wards  Supplementary appendix.docx