Breastfeeding and perceptions of breast shape changes in Australian and Japanese women

Abstract

This thesis examines infant feeding practices, including knowledge and attitudes towards breastfeeding, factors that influence the duration of breastfeeding, and breastfeeding outcomes in relation to postpartum women’s body image, particularly perceived changes of breast shape in Australian and Japanese women.From previous studies, the number of women who desire to be slim after their childbirth has been increasing in recent years, and it is hypothesised that their concerns about body image may be associated with breastfeeding outcomes. It has been found that the Asian women have more distorted body images than their Caucasian counterparts. Breastfeeding gives many benefits to health of women and child, and the benefits may be greater with longer breastfeeding duration. Nevertheless, the duration of ‘exclusive breastfeeding’ is still less than the recommendations in most countries. There have been only a limited number of studies investigating breastfeeding outcomes related to postpartum women’s body image.This thesis contains two studies. In the first study, a cross-sectional survey was conducted in 2007 at Himeji city, Japan, in order to describe infant feeding practices and to identify factors that are associated with breastfeeding duration. The survey was a self-administered questionnaire included in a letter from the Himeji City Health Department offering free health checks for infants at 18 months of age. The second study was a prospective cohort study with interviews on seven occasions (4, 10, 16, 22, 32, 40, and 52 weeks postpartum) of follow-up over a period of 12 months, the Perth Infant Feeding Study Mark II (PIFS II). This second study was undertaken in two local hospitals in Perth, Australia between 2002 and 2003. While the initial results of the PIFS II have been published elsewhere, variables related to the postpartum women’s body image had not previously been analysed. Both studies used almost the same questions on infant feeding practices and about postpartum women’s body image.Univariate analysis was first employed to describe breastfeeding practices and to assess which factors could be included in the model of breastfeeding duration. These were then incorporated into multivariate analysis to identify the contributing factors of the duration of ‘any breastfeeding’ to six months postpartum. In the second group of analysis, factors related to postpartum women’s body image, their Body Mass Index (BMI) and perceptions of changes in their weight status, body and breast shape were included. These factors were then entered into a multivariate logistics regression analysis with other contributing factors in breastfeeding duration.In the Japanese study, 2,345 women were eligible to participate and 1,623 completed the questionnaire (Response rate: 69.2%). For the final analysis, 1,612 questionnaires were used, due to missing data. The majority (72.8%) of respondents were housewives with mean age 32.2, ranging from 18 to 44. Almost half of them were primiparous (56%) and had annual family income below 4,500,000 Japanese yen (JP¥) (55.5%) (Equivalent value in Australian dollars (AU$), approximately 50000). Low birth weight infants (less than 2500g) were 8.4%, and the great majority of postpartum women (85.0%) had a vaginal delivery. The prevalence of smoking and alcohol during breastfeeding reported by the women were 10.6% and 14.1%, respectively. It was interesting to note that the first feed for 29.8% of the infants was infant formula, followed by glucose (19.5%). By two weeks of age, more than 85% of infants had been introduced to infant formula. Consequently, ‘exclusive breastfeeding’ was not common and ‘any breastfeeding’ was the most common infant feeding method in Japan, with mean duration of 9.1 months (Standard Deviation (SD) = 6.4). The Iowa Infant Feeding Attitudes Scales (IIFAS), reflecting knowledge and attitudes towards infant feeding practices, of the sample was 54.2 (SD = 4.9), a low score, compared with other studies. In this sample, other family members such as husbands and women’s own mothers were less involved in decisions on infant feeding methods.With regard to postpartum women’s body image, 23.9% were in the ‘underweight’ category in their Body Mass Index (BMI) before the first pregnancy and this decreased to 17.8% at 18 months postpartum. The majority perceived some changes in their breast shape, namely ‘lost tension’, ‘changes in size’, and ‘sagging’. While more than half believed that breastfeeding helped to regain their body shape faster than infant formula, the same proportion stated that breastfeeding also made their breast shape ‘drooping’. The Attention to Body Shape Scale (ABS), a measure assessing one’s body image, showed that the Japanese postpartum women were very conscious of their body shape, but this was not a significant factor in relation to breastfeeding outcomes.Results of stepwise logistic regression analysis found four factors that were negatively associated with the duration of ‘any breastfeeding’ to six months of postpartum specifically, working women in non-professional areas (adjusted Odds Ratio (OR) = 0.41, 95% Confidence Interval (CI) = 0.23 to 0.72), experienced in breastfeeding problems (adjusted OR = 0.34, 95% CI = 0.23 to 0.51), smoking by the women during breastfeeding (adjusted OR = 0.32, 95% CI = 0.15 to 0.66), and perceived changes in breast shape (adjusted OR = 0.60, 95% CI = 0.40 to 0.91). In contrast, four factors, the later introduction of infant formula (adjusted OR = 1.07, 95% CI = 1.03 to 1.11), primipara (adjusted OR = 2.26, 95% CI = 1.05 to 4.87), the decision on infant feeding methods before pregnancy (adjusted OR = 2.73, 95% CI = 1.79 to 4.18), alcohol intake during breastfeeding (adjusted OR = 2.00, 95% CI = 1.18 to 3.39) were significantly associated with continuing ‘any breastfeeding’ to six months postpartum.In the Australian study, the Perth Infant Feeding Study Mark II (PIFS II), overall 870 of 1068 postpartum women who were eligible for the study were recruited and asked to complete the baseline questionnaire. In total, 587 completed questionnaire were received, a response rate of 68%. Approximately 30% of respondents were in their 30s, which was the highest proportion by age. The majority were Australian born (73.2%) and were multiparous (63.2%). Only 2.2% of infants were low birth weight and 70% had vaginal deliveries. The highest percentage of annual family income was at the group between 25,000 AU$ and 40,000 AU$ at 31.0%. The smoking rate reported by the women lies between 23% and 39% during the follow-up periods, and the prevalence of alcohol consumption by the women ranged from 40.2% to 48.5%. The majority had made a decision about infant feeding methods either before or early pregnancy (91.8%) and those infants were most often fed breastmilk as their first food while in hospitals (81.2%).Their husbands and own mothers positively supported breastfeeding in the decision making processes. The mean duration of ‘any breastfeeding’ and of introducing infant formula were 23.3 weeks (SD = 19.7) (approximately 5.8 months). The IIFAS also indicated that the Australian women showed positive attitudes towards breastfeeding with the mean score 64.4 (SD=8.3). About 8% of the Perth women were in the ‘underweight’ category (BMI < 18.5) before pregnancy. 51% noticed changes in their breast shape within four weeks of birth and this increased to 60% at 52 weeks postpartum. The most common changes perceived between four and 52 weeks postpartum were ‘sagging’ following ‘changes in size’. The factors that influenced breastfeeding duration in Australia were similar to the Japanese women with the exception of alcohol consumption. A multiple logistic regression showed that the women who had longer breastfeeding duration were more likely to perceive changes of their breast shape at week 52. For a one week increase in the duration of ‘exclusive breastfeeding’ up to six months, the average risk of perceived changes in breast shape at week 52 increased by 3.5% (95% CI = 1.00 to 1.07), by 3.8% (95% CI = 1.01 to 1.07) with ‘any breastfeeding’ duration up to six months, and by 2.9% (95% CI = 1.00 to 1.06) with ‘full breastfeeding’ up to six months, respectively. Moreover, the average risk of perceived changes in breast shape at week 52 increased by 0.8% (95% CI = 1.00 to 1.02), whenever women increase ‘any breastfeeding’ duration up to 12 months with each additional week.Cox logistic regression was also used to explore breastfeeding by type for six months or 12 months given their perceptions of breast shape at four weeks postpartum. However, no significant associations were found after adjustment of the model. The results indicated that the Australian women perceived changes to their body weight and breast shape in the postpartum period, but Japanese women were significantly more conscious about these aspects than their Australian counterparts. The cultural backgrounds of the women affected the associations between body image and breastfeeding outcomes. It is important to acknowledge this factor when providing education programs in breastfeeding.Breastfeeding outcomes were different between studies, particularly with regard to supplemental feeding, family involvement in decisions about infant feeding methods, and opinions on alcohol intake during breastfeeding. In addition, the prevalence of women with ‘underweight’ BMI is different between the two countries. These two studies suggest that cultural beliefs about women’s body image were associated with breastfeeding duration. The Japanese women were more conscious about their body and breast shape, which was significantly associated with breastfeeding duration, compared with their Australian counterparts. Women’s knowledge and attitudes towards breastfeeding reflected this and was at a lower level in Japan than in Australia. Several issues including media exposure, social expectations about women’s body shape and gestational weight control that are potentially related to women’s body image are discussed below.There are some limitations that needed to be considered when interpreting and assessing the applicability of the results. Through these two studies, changes of body weight, breast in shape were not assessed by any objective measure and were based on the participants’ perceptions. In addition, as the first study was cross-sectional and based on self-reported data at 18 months postpartum in Japan, there is a possibility of some recall bias about breastfeeding practices. In the second study, the PIFS II, a limitation was that the sample was recruited from two public hospitals and probably reflected a selection bias towards lower socio-economic status.It is recommended further studies using a cohort method with a large sample size and a qualitative approach be undertaken to enrich and verify the results of the study. These will enable a more accurate estimate of the prevalence of breastfeeding, factors that predict breastfeeding duration and postpartum women’s perceptions of changes in breast shape during breastfeeding. Breastfeeding education programs should be based on accurate studies of maternal knowledge and attitudes towards infant feeding methods, benefits of breastfeeding in the short and long term, and attitudes to body image in postpartum. These education programs need to acknowledge cultural perspectives in breastfeeding and women’s body image. Further improvements in working environments of postpartum women allowing them to continue breastfeeding are also suggested. It is important for public health that relevant interventions should be developed to improve breastfeeding rates and duration, in Australia and Japan.