While breastfeeding initiation within one hour of birth has increased in Cambodia since 2000 [21], additional efforts to facilitate timely initiation should be focused on women with noted barriers. Among women who reside in poor households, the likelihood of timely breastfeeding initiation was lower in women who gave birth at home compared to those who gave birth in public health facilities. Urban women giving birth in private facilities were less likely to timely initiate breastfeeding than urban women giving birth in public facilities, and women who gave birth in private facilities had greater rates of offering prelacteal feeds than women who gave birth in public facilities. These results suggest public health facilities and their health professionals are more likely to promote breastfeeding initiation than private facilities and home birth attendants.

A cross-sectional study found that 45.1% of mothers in Phnom Penh reported seeing breastmilk substitute advertisements in health facilities [20]. Violations of the Code are reported among private health facilities [28], as no monitoring system was implemented, and no penalties or fines were imposed on violators of the Code in Cambodia [29]. The 2015 Cambodia World Breastfeeding Trends Initiative (WBTi) reports the most popular private maternity facility distributes free infant formula [29]. Due to the weak monitoring of infant formula companies and their health facility partners, formula promotion violating the Code goes unrecognized and unreported [29]. The Cambodia WBTi designated only 47% of hospitals as Baby-friendly in the past five years, all of which were public health facilities [29], further depicting the influence place of birth can have on timely breastfeeding initiation.

As private facilities are concentrated in urban areas [32] and Phnom Penh is the most populous, urban city in Cambodia, women giving birth in urban private settings were possibly more exposed to breastmilk substitutes from infant formula companies and their health professional partners [28, 29], than women giving birth in rural private facilities. This trend may also contribute to the lack of BFHI buy-in in the private facilities [28] and to the discrepancy in timely breastfeeding initiation. Even though Cambodia’s commitment to universal healthcare has made recent progress, the private sector is skewed towards the richest quintile [25], while rural health centers tend to service the poor [32]. The Cambodian Ministry of Health has appropriated more medical staff, mainly nurses and midwives, to health centers in rural provinces and remote areas which had faced healthcare disparities due to low salaries and intense work conditions [32]. Facilitating partnerships among private and public sectors will grant more equal access to quality healthcare [25, 32, 33], as well as communicating current evidence-based breastfeeding care and programs.

These findings correspond with the increasing trend of breastmilk substitute use and its marketing, especially among women who give birth in private facilities. Previous findings show that 26.1% of women who gave birth in private facilities provided their infants younger than 6 months of age with breastmilk substitutes, five times more than women who gave birth in public facilities [21]. No association was found among rural women giving birth in private facilities on timely breastfeeding initiation, suggesting rural women may have a lower exposure to breastmilk substitutes compared to urban women in private facilities, thereby reducing their risk of impeding timely initiation by offering breastmilk substitutes. Rural women giving birth at home were less likely to timely initiate breastfeeding than urban women giving birth at home. This may be due to education or local community practices and beliefs [34], and family influences, as many women reported their mother or sisters were their main sources of breastfeeding advice at home [35]. In contrast, there was no significant association among urban women giving birth at home and timely initiation. As ‘urban’ indicates a modern, city setting, women residing in urban areas may forgo traditional postpartum practices.

Midwives and traditional birth attendants typically assist with home births [36], and in Cambodia, midwives are especially utilized in health centers [32]. Because of their role in antenatal care, midwives’ attitudes and actions can influence new mothers’ decisions to timely initiate breastfeeding. Midwives reported encouraging mothers to breastfeed after birth, however, timely breastfeeding was not observed in the first hour postpartum which may be explained by maternal feelings of modesty in the presence of an observer and a sense of discomfort discussing breastfeeding by younger midwives [35]. To encourage midwives and TBAs, Cambodia launched the Government Midwifery Incentive Scheme (GMIS) in 2015, providing cash incentives for midwives and other trained health personnel per live birth they attended in health facilities, with higher pay for health center births in rural areas [37]. GMIS helped decrease maternal mortality rates from 473 in 2005 to 206 in 2010 by increasing health facility births [19, 37]. Government Midwifery Incentive Scheme success in Cambodia, proposes performance-based incentive programs may also be an effective tool in improving related indicators of breastfeeding initiation, such as prolonged SSC following birth, assisting with timely breastfeeding initiation, and providing lactation support to new mothers. Reproductive, maternal, newborn and child health (RMNCH) programs [38], like EENC and GMIS, aim to improve healthcare financing and workforce in Cambodia, including universal healthcare coverage, increasing staff, and providing cash incentives [25, 32].

From 2000 to 2014, public health facility births increased in Cambodia by 82 percent [23]. The growing trend of exclusivity and timely breastfeeding initiation in Cambodia may be correlated with public health facilities that adopted the WHO/ UNICEF BFHI [22]. However, as previously mentioned, many countries have halted BFHI programs [28], or have merged or adopted the EENC Action Plan [14]. Cambodia is one of two countries achieving the Action Plan target of 80% of facilities introducing EENC, including immediate and sustained skin-to-skin contact [18]. Over half of hospitals reported forming EENC teams, however, only 19% reported conducting essential, routine quality care assessments, suggesting the need to gain health professionals’ and facilities’ support for the program.

The findings of the current study highlight the importance of educating and incentivizing health professionals to promote and regularly assess breastfeeding support, specifically those assisting private facility or home births. Support of timely initiation and lactation guidance extends beyond verbal encouragement [2, 35], therefore, promotion of EENC, for example, can improve timely initiation rates, breastfeeding exclusivity, and duration in Cambodia [18].

Women with cesarean births had lower odds of timely breastfeeding initiation regardless of household wealth index, with lower odds of timely breastfeeding initiation, compared to women who gave birth vaginally (65% among poor, 87% among middle, and 74% among rich). Efforts to support women post-cesarean should be implemented across all socioeconomic groups. Results are corroborated by previous research finding decreased odds of timely breastfeeding initiation among women who had cesarean births in low- and middle-income countries [39, 40]. Delayed initiation may be due to post-cesarean pain or fatigue, difficulty with breastfeeding positioning related to the incision site, and maternal-infant separation [39, 41, 42].

Multiparous women and those who reside in middle wealth index households were more likely to timely initiate breastfeeding. Multiparous initiation rates are consistent with other southeast Asia studies suggesting women have greater self-efficacy based on previous breastfeeding experience, contributing to timely initiation [40, 43].

Other factors significantly associated with timely breastfeeding initiation included antenatal visits, marital status, type of birth attendant, and prelacteal feeds. Greater odds of timely breastfeeding initiation were observed among women residing in the middle wealth index; however, limited research exists to explain the correlation of household income and timely breastfeeding initiation. A higher number of antenatal care visits (1–3 and 4 +) were significantly associated with timely breastfeeding initiation among urban women in the stratified analysis. Similar findings were noted in India and Ethiopia, suggesting greater antenatal care visits provide more education and support for early breastfeeding [44, 45]. Married women and those receiving ‘husband support’ may be more likely to timely and exclusively breastfeed due to perceived additional support from their partner [46, 47]. In our study, urban women had greater initiation rates when assisted by a traditional birth attendant, suggesting that their services promote breastfeeding and reduce costs compared to physician or midwife services [48].

Employed women had a lower likelihood of timely breastfeeding initiation, consistent with a systematic literature review conducted in South Asia [49], suggesting women may not feel supported in breastfeeding as they intend to return to work. One of the largest industries in Cambodia, employing mostly young women, is the garment industry [50]. Maternity protection laws in Cambodia grant 90 days for maternity leave, affords one hour per day is paid time off for breastfeeding, and upon returning to work, specifies that duties are lighter for two months [50]. Most women report receiving paid time off to breastfeed, however, most women lack a childcare center at or near the factory [50]. Cambodian law states that factories employing over 100 women must provide a nursing room, however compliance is low, and women report abstaining from bringing their infants to work [50]. Employer incentives could be offered to companies complying with laws protecting maternity rights.

Women offering prelacteal feeds had greater odds of not practicing timely breastfeeding initiation regardless of household wealth index and place of residence. Women residing in rural areas who offered prelacteal feeds had the lowest odds of timely breastfeeding initiation. Although national public health breastfeeding campaigns have helped stabilize breastmilk substitute use with no increase since 2005, limited resources inhibit the continuation of communication campaigns [21]. In the absence of such campaigns, breastmilk substitute companies may increase illegal promotion among vulnerable populations [21], such as those of rural settings without regular media exposure.

Choosing to offer prelacteal feeds, may also be explained by socio-cultural differences between urban and rural areas when considering characteristics and influences of the individual, her household, and community [51]. In May 2014, a press release published in Cambodia called for improved commercial adherence to the WHO and UNICEF International Code of Marketing of Breast-Milk Substitutes and Cambodia’s Sub-Decree 133 [52], whereby the benefits and superiority of breastfeeding must be clearly explained and advocated [53, 54]. This press release cited 113 different breastmilk substitutes being marketed in Phnom Penh with none of them fully compliance with Sub-Decree 133 [52]. In urban settings, such as Phnom Penh, breastmilk substitutes or prelacteal feeds, includes infant/starter formula (indicated for birth to five months of age) [17], while the CDHS includes breastmilk substitutes as formula, sweetened condensed milk, other canned milk usually thinned with water, or watery rice porridge (borbor) [19]. Differences in type of breastmilk substitutes offered is not discernable between rural and urban populations of Cambodia. However, the extensive availability and variety of breastmilk substitutes may contribute to their use in rural and private health centers, as the private health sector continues to market them aggressively [21].

The strength of this study is the large, representative sample of Cambodian women and the availability of various potential confounders for adjustment in the multivariable models. Additionally, all data were collected using a well-establish, standardized and rigorous methodology implemented by the DHS program. Limitations of this study include the cross-sectional design preventing the determination of causation. The role of various skilled birth attendants and health professionals regarding breastfeeding was not captured; therefore, we were unable to determine the extent of breastfeeding counseling and support in the prenatal and early postpartum periods. Additionally, results may have been influenced by self-report and social desirability biases. Lastly, the data lacked information regarding the administration of prelacteal feeds specifically in the first hour, precluding determination of the impact of prelacteal feeds on breastfeeding in the first hour after birth.