Breast-feeding in Athens, Greece

Factors Associated With Its Initiation and Duration

Abstract

Objectives: 

To determine the prevalence, examine the influence of hospital practices and investigate potential determinants of breast-feeding in Athens.

Patients and Methods: 

Three hundred twelve mothers provided information regarding feeding practices at certain maternity hospitals in Athens, at 40 days and 6 months postpartum. Multiple logistic regression analysis was performed to evaluate the association between the initiation and maintenance of breast-feeding and potential risk factors.

Results: 

Although almost 90% of newborn infants were given a breast milk substitute one or more times during the first 2 days at the maternity hospital, the exclusive breast-feeding percentage on the last day of hospital stay reached 85%. Breast-feeding and exclusive breast-feeding percentages dropped to 55% and 35%, respectively, at 40 days postpartum and to 16% and 12%, respectively, at 6 months postpartum. While in the hospital, 3% of mothers initiated breast-feeding within 1 hour of labor, only 34% were informed about the advantages of breast-feeding by health professionals and 42% were trained to breast-feed by the midwives. "Rooming-in" was not practiced in the private hospitals. The educational level was positively associated with the initiation of breast-feeding [odds ratio (OR):1.36, confidence interval (CI):1.02-1.81], the mother's body mass index was negatively associated with the maintenance of breast-feeding for 40 days (OR:0.56, CI:0.32-0.98) and 6 months (OR:0.28, CI:0.06-1.26) and a caesarean section was negatively associated with the initiation (OR:0.24, CI:0.11-0.49) and maintenance of breast-feeding (OR:0.42, CI:0.20-0.89).

Conclusions: 

Breast-feeding is not appropriately supported in certain maternity hospitals in Athens, and this is probably the cause of observed low breast-feeding prevalence.

INTRODUCTION

Within the last half-century, there have been major changes worldwide in the type of feeding predominantly offered to infants before the introduction of complementary foods, involving a switch from breast-feeding to formula milk feeding based on cow's milk. After reviewing the advantages of human milk, the American Academy of Pediatrics (¹) and the World Health Organization (WHO) (²) strongly adhere to the belief that breast-feeding is the optimal infant feeding practice. Furthermore, the WHO recommends exclusive breast-feeding for 6 months and continued breast-feeding for up to 2 years and beyond (³).

Although there is overall recognition that human milk ensures the best possible outcomes for the infant, in many countries breast-feeding rates are still unsatisfactory. In the United States, as indicated by Ryan et al. (⁴), "Breast-feeding continues to increase into the new millennium"; but as stated by Li et al. (⁵), breast-feeding and exclusive breast-feeding rates are still low. In Europe, breast-feeding rates vary over a wide range (⁶). The rate of exclusive breast-feeding at 6 months is low almost everywhere, even in countries with high initial rates (⁷).

The present study aims to provide updated rates of breast-feeding for up to 6 months postpartum to investigate potential determinants for the initiation and duration of breast-feeding and to examine the influence of hospital practices on breast-feeding prevalence in Athens, Greece.

PATIENTS AND METHODS

This prospective study presents the results of an epidemiological survey conducted from September 2003 to February 2004 in Athens, Greece, that included a total of 312 participants. In Greece, a large percentage of deliveries (approximately 50%) take place in private maternity hospitals. Furthermore, in Athens, mothers belonging to a middle-to-low socioeconomic group usually deliver in public hospitals, whereas mothers belonging to a middle-to-high socioeconomic group usually deliver in private maternity hospitals. Therefore, to recruit a representative sample of women for the study, participants included mothers who delivered in public and in private maternity hospitals situated in the center of Athens.

Two hundred eight women who consented to participate from a total of 280 consecutive women who delivered in the largest public university maternity hospital in Athens within a 2-month predetermined period of time (September and October 2003) were recruited into the study on the last day of their hospital stay. The study was approved by the scientific board of the public university maternity hospital. As no approval could be requested from the administrative boards of the private maternity hospitals to conduct the present study, 104 consecutive women who had babies younger than 3 months old and who attended 3 private pediatric clinics or the outpatient department of a children's university hospital, also within a 2-month predetermined period of time (January and February 2004), were also included in the study. These mothers were required to have babies younger than 3 months old (to include mothers who delivered within the last 3 months of 2003) to help them accurately recall the facts that were required to answer the simple questions listed in the questionnaire. Thus, a total of 312 women entered the study.

Although the sample size was not predetermined, it was a posteriori ascertained that the sample size of 312 mothers was adequate to evaluate differences in the odds ratios (ORs) greater or equal to 20% at a P value less than 0.05, giving a statistical power of 80% assuming a simple random sampling procedure.

All mothers were asked to answer the questions of an extended questionnaire (>40 questions) during a personal interview. The information collected included sociodemographic and anthropometric data, information about race and ethnicity, maternal education, smoking habits, maternal illnesses, socioeconomic status (determined by each mother's insurance cover, profession and current home address), health status of the newborn, questions related to each mother's desire to breast-feed, their previous experience of breast-feeding, infant postpartum feeding information at the maternity hospital (recall period since birth) and information on hospital practices regarding breast-feeding support (recall period since birth).

At the public university maternity hospital, the duration of stay was usually 3 days for mothers who delivered naturally and at least 4 days for mothers who delivered by caesarean section.

Those mothers who delivered in the public university maternity hospital and indicated their intention to continue breast-feeding consented to further contacts by phone 40 days postpartum. They were asked to provide information relating to the type of milk fed to their infant within a 24-hour recall period; and when applicable, they were also asked the reasons for breast-feeding cessation.

Apart from answering the questions posed in the questionnaire, those mothers who were recruited while attending the private clinics of 3 pediatricians or the outpatient department of the Children's University Hospital were asked in their initial interview to recall the type of milk fed to their infant on the 40thday postpartum and the reasons for breast-feeding cessation when applicable.

All participants who breast-fed for 40 days postpartum were also contacted at 6 months and were asked to recall the type of milk fed to their infant within the last 24 hours. When applicable, reasons for breast-feeding cessation were recorded. All data were entered in the SPSS program (version 10.0) for further analysis.

To have a common set of measures to assess breast-feeding practices, 2 categories of breast-feeding practices were considered:breast-feeding and exclusive breast-feeding. According to the definitions recommended by WHO (⁸), the term breast-feeding is used when the child has received breast milk (direct from the breast or expressed) and the term exclusive breast-feeding is used when the infant has received only breast milk from his or her mother or a wet nurse, or expressed breast milk and no other liquids or solids, with the exception of drops or syrups consisting of vitamins, mineral supplements or medicines.

Statistical Analysis

In the present study, quantitative variables are presented as mean ± SD, whereas categorical variables are presented as absolute and relative frequencies. Associations between categorical variables were tested by the use of contingency tables and the calculation of χ² tests. The relative probability to initiate and maintain breast-feeding was calculated by estimating the ORs and their corresponding 95% confidence intervals (CIs) through multiple logistic regression analysis after adjustment for several potential confounders and interactions.

The final risk model was developed through a backward elimination procedure for the selection of variables based on the Wald statistic and using a 5% probability for entry and 10% probability for removal of a variable from the model. Deviance residuals were calculated to evaluate the model's goodness-of-fit.

RESULTS

A total of 312 mothers participated in the study. Two hundred twenty-nine (73%) of them delivered in a public maternity hospital, 82 (26%) delivered in a private maternity clinic and 1 mother did not answer the question.

Of the 229 (73%) mothers who delivered in a public hospital, 208 deliveries (67%) took place at the university public hospital, whereas the remaining 21 mothers delivered in other public maternity hospitals situated in Athens. Personal, sociodemographic, perinatal and postnatal characteristics for the mothers are shown in Table 1.

TABLE 1: Personal, sociodemographic, perinatal and postnatal characteristics for the mothers in Athens, Greece

All illiterate, black, Asian and Gypsy mothers living in a low-economic-level area delivered at the university public hospital. In addition, 82 of the 88 foreign women delivered at the same hospital. Most foreign mothers stated that they were immigrants for economic reasons.

Most mothers (269; 86%) were healthy, although a small percentage reported that they suffered from hypertension (5; 2%), hyper/hypothyroidism (15; 5%) and miscellaneous ailments (22; 7%); and 1 (0.3%) gave no answer.

The mean ± SD weight of the mothers before pregnancy and before delivery was 62 ± 11 kg and 76 ± 12 kg, respectively, accounting for an increase of 14 ± 8 kg in body weight during pregnancy. The mean body mass index (BMI) of the mothers before pregnancy was 22.4 ± 5.3.

Only 34% of the mothers were informed about the advantages of breast-feeding by health professionals (10% from their gynecologist, 9% from the pediatrician and 15% from the maternity nursing staff), whereas 29% were not given any kind of information. In addition, only 42% of the mothers were trained to breast-feed by the maternity nursing staff. The answers of the mothers regarding feeding practices and hospital practices at the maternity hospital, as well as breast-feeding duration and the reasons for breast-feeding cessation, are recorded in Table 2.

TABLE 2: Hospital practices on BF support, BF duration and reasons for BF cessation

Although almost 90% of newborn infants were given a breast milk substitute one or more times on the first day and sometimes the second day postpartum, 85% breast-fed exclusively on the last day of their stay in the maternity clinic.

It should be noted that only 1 mother who delivered in a private maternity hospital reported that she remained together with her newborn infant 24 hours per day ("rooming-in" practice).

Based on the results of multivariate logistic regression analysis, mothers with an academic education had a higher probability to initiate breast-feeding within 1 hour of birth (interaction not statistically significant because of the small number of mothers who initiated breast-feeding within 1 hour of labor) and a 35.7% higher probability to initiate breast-feeding within 24 hours of labor.

In addition, mothers who delivered by caesarean section had a 76.3% lower probability to initiate breast-feeding within 24 hours of labor and a 57.9% lower probability to maintain breast-feeding for 40 days compared with mothers who delivered naturally.

Finally, mothers with a high BMI before pregnancy had a 43.8% lower probability to maintain breast-feeding for 40 days and a 72.4% lower probability to breast-feed for 6months compared with mothers with a low BMI. Further analysis showed that overweight (≥25 kg/m²) women were 43% less likely to breast-feed at 40 days postpartum (OR:0.57, CI:0.34-0.91, P < 0.05), whereas obese (≥30 kg/m²) women were 24% less likely to breast-feed (OR:0.76, CI:0.54-0.99, P < 0.05). In addition, overweight and obese women were 50% less likely to breast-feed at 6 months postpartum (OR:0.50, CI:0.28-0.90, P < 0.01). No association was found between smoking and the initiation and duration of breast-feeding. The individual effect of the educational level, the mode of birth and the BMI is presented in Table 3.

TABLE 3: Results from the multivariate logistic model that was developed for the evaluation of potential risk factors on the initiation of breast-feeding within 1 hour and 24 hours after labor and on the maintenance of breast-feeding for 40 days and 6 months postpartum

DISCUSSION

In the maternity hospitals studied, breast-feeding rates on the last day of hospital stay were found to be quite high (85%) compared with breast-feeding initiation rates in the United States (^4,5) and Europe (⁶), despite the fact that almost 90% of newborn infants had received at least one bottle of formula while at the maternity hospital. However, these rates subsequently dropped, suggesting that breast-feeding rates in Athens, judging from the results of the study, are far from the goal of the WHO, which promotes exclusive breast-feeding for 6 months (³). More specifically, 55% of the mothers breast-fed and 35% breast-fed exclusively for 40 days. Furthermore, 16% breast-fed for 6 months but only 12% breast-fed exclusively. Antoniou et al. (⁹) recently published a study on breast-feeding rates in Greece in 2001 and reported a similar breast-feeding percentage at discharge. However, breast-feeding percentages at 6 months were lower in our study, indicating that there was no improvement between 2001 and 2003 (⁹).

Among the factors thought to influence breast-feeding initiation and duration was the calculated caesarean delivery rate that was found to be 38%, an exceedingly high rate considering the 2003 caesarean delivery rate in the United States, which was 27.6% (¹⁰). The caesarean section does not support the initiation or the maintenance of breast-feeding (^11,12).

In addition, we found that mothers with a high educational level were more likely to breast-feed compared with illiterate mothers. This is in agreement with the findings of a literature review conducted by Callen and Pinelli (¹³) and with studies conducted by Bertini et al. (¹⁴) and Ummarino et al. (¹⁵). However, in our study no association was found between the educational level and the duration of breast-feeding.

A relationship between the mothers' BMI and the prevalence of breast-feeding was also found. Overweight and obese mothers had a smaller probability to maintain breast-feeding for 40 days and 6 months after labor compared with normal-weight mothers. This result is consistent with data from studies conducted by Hilson et al. (¹⁶) and Li et al. (¹⁷). Current literature also shows that maternal prepregnant BMI is associated with infant weight gain (¹⁸).

Although many studies have been published in the international literature confirming that maternal smoking affects breast-feeding prevalence (^11,19), no interactions were found in the present study between maternal smoking habits and the initiation and duration of breast-feeding, probably because of the small number of mothers who were smoking during pregnancy. The high percentage of nonsmoking mothers in this study (81%) suggests a significant degree of compliance with the guidelines that encourage pregnant mothers to refrain from smoking.

Furthermore, although modern society promotes women's breasts as a sex symbol instead of supporting their feeding role, none of the mothers in this study reported that they stopped breast-feeding for aesthetic reasons- an encouraging finding.

To promote breast-feeding, in 1995 the Greek government distributed information leaflets pertaining to "Baby Friendly Hospitals" and the "Ten Steps to Successful Breast-feeding"(²⁰). Although this initiative (²⁰) resulted in increasing significantly breast-feeding rates in many countries (^21-25), information regarding compliance with these guidelines is not available in Greece.

Only 10% of the participants reported that they exclusively breast-fed their infant in the maternity clinic during the first 2 days, whereas almost 90% of mothers reported that milk formula was administered to their baby one or more times. The lack of breast-feeding hospital practices in public and private maternity hospitals as well as the absence of the rooming-in practice in the private maternity hospitals probably contributed to this finding. From the data collected in this study, the effect of rooming-in alone could not be ascertained because hospital practices in general do not support breast-feeding. In addition, the overwhelming majority of mothers interviewed were not assisted to initiate breast-feeding within the first half-hour after childbirth. Only 9 mothers (3%) reported that they initiated breast-feeding within 1 hour after delivery.

Two years ago, Antoniou et al. (⁹) reported a similar percentage of breast-fed infants within the first hour after labor but a lower percentage of infants given formula in the maternity hospital (47.1% versus 89% in our study), indicating a deterioration instead of an improvement in hospital practices with respect to breast-feeding support.

Furthermore, it is remarkable that only 34% of the mothers were given any information about the advantages of breast-feeding by health professionals and that only 42% were given instructions on breast-feeding technique by the maternity nursing staff, indicating a limited degree of interest from responsible health care professionals. Although it is a natural act, breast-feeding is also a learned behavior, and virtually all mothers can breast-feed provided they have accurate information and support. Labarere et al. (²⁶) demonstrated that breast-feeding support does contribute to improved breast-feeding prevalence and suggested that training programs for practicing physicians would probably help. In contrast, de Oliveira et al. (²⁷) and Guise et al. (²⁸) cast doubts about the effectiveness of health education alone.

The analysis of the data collected in our study suggests that inadequate hospital practices, and in particular the almost universally accepted routine of giving a formula supplement to healthy newborn infants, thus causing the mothers to believe that their milk supply may be insufficient, are responsible for the unacceptably low rate of breast-feeding and exclusive breast-feeding among infants during the first 6 months of life. Such low rates occur despite the evident desire of the mothers to breast-feed, as shown by the high initiation rate found in the study.

CONCLUSIONS

The results of this study indicate that although mothers intend to breast-feed when leaving hospital, breast-feeding prevalence is quite low and this may be because of lack of appropriate support for breast-feeding in certain maternity hospitals in Athens.

Breast-feeding is the optimal method of feeding infants and young children, reduces the risk of some common childhood illnesses and is a basic component of child and adult health. Thus, low breast-feeding rates should be recognized as a public health problem and should be managed as such. A renewed commitment to promote the health and nutrition of infants in all of the maternity hospitals in Athens would be advisable.