نگرش تغذیه پس از زایمان، افسردگی مادر، تغذیه با شیر مادر و در باربادوس

Abstract Maternal feeding attitudes, maternal moods and infant feeding practices during the first 6 months postpartum were assessed in 226 healthy, well-nourished Barbadian mother–infant dyads. Factor analysis of the feeding attitudes questionnaire resulted in six independent factors. The belief that breastfeeding was better than bottle-feeding was associated with higher family income, more information seeking behavior and older maternal age at the time of her first pregnancy. Women who believed that breastfeeding was better at 7 weeks postpartum were also more likely to breastfeed at concurrent and later ages, up to 6 months postpartum. This belief was also associated with less maternal depression at 7 weeks and 6 months. The association between feeding attitudes and actual feeding practices was significant even after correcting for maternal moods and other background variables. Conversely, after controlling for feeding attitudes, maternal mood at 7 weeks was still significantly associated with infant feeding practices at 6 months. Thus, feeding attitudes and maternal moods were closely linked, but each contributed independently and uniquely to different aspects of breastfeeding, especially at 6 months. These findings suggest that early intervention addressing maternal feeding attitudes, may improve the extent of breastfeeding and the health of children in this setting.

Introduction Although breastfeeding has been recognized as the optimal method of infant feeding for the first 6 months of life, many mothers in both developed and developing countries do not breastfeed (American Academy of Pediatrics Work Group on Breastfeeding, 1997; UNICEF, n.d. and WHO, 2000). Many studies concerning the determinants of breastfeeding are focused on the assessments of socioeconomic and demographic factors, in addition to medical histories (Ryan, Wenjun, & Acosta, 2002). However, feeding attitudes have been recognized as being stronger predictors of the choice and duration of breastfeeding than demographic factors (Baisch, Fox, Whitten, & Pajewski, 1989; Hill, 1988; Scott, Shaker, & Reid, 2004). Thus, investigators considering the psychosocial aspects of breastfeeding and reasons for its early termination among low-income women, in particular, have recognized the important role of maternal perceptions and views about the benefits and social acceptance of breastfeeding (Baranowski, Rassin, Richardson, Brown, & Bee, 1986; Guttman & Zimmerman, 2000). In a series of studies in low-income U.S. women, feeding attitudes were found to predict breastfeeding intention, irrespective of parity or prior breastfeeding experience (Kloeben-Tarver, Thompson, & Miner, 2002). In Western countries, the literature has generally confirmed that feeding intention is closely associated with actual feeding practices (Caulfield, Gross, & Bentley, 1998; Kessler, Gielen, & Diener-West, 1995). There are few studies with detailed information regarding the relationship between feeding attitudes and their impact on breastfeeding in other cultures, where breastfeeding may be even more important to the health and well-being of infants in societies where economic and medical resources are less readily available. A recent survey of feeding attitudes among South African women reports that health benefits to the infant was the major reason for breastfeeding (Sibeko, Dhansay, Charlton, Johns, & Gray-Donald, 2005). The purpose of the current study, performed in Barbados, West Indies, was to document feeding attitudes at 7 weeks postpartum and concurrent and later feeding practices in a population of healthy mothers and infants over the first 6 months of life. In the early 1980's, Barbados had been recognized by UNICEF as having one of the lowest rates of breastfeeding in a developing country (UNICEF, 1980). Determining psychosocial factors which contributed to breastfeeding success and improved infant outcomes on this island was the major aim of our series of studies. In previous papers, we described the association of infant feeding practices and postpartum maternal moods on the growth and development of 226 well-nourished children during their first 6 months of life and at 11 years of age. Breastfeeding positively impacted the physical growth of Barbadian children (Galler, Ramsey, Harrison, Brooks, & Weiskopf-Bock, 1998). Although breastfeeding per se was not associated with performance on the Griffiths Developmental Scale (Griffiths, 1954) in the first 6 months of life (Galler, Harrison, Ramsey, Forde, & Butler, 2000), infant length, which was closely associated with the amount of breastfeeding, was significantly correlated with cognitive performance at 11 years (Galler, Ramsey, et al., 2004). In contrast, maternal depressive symptoms were significantly and independently associated with reduced breastfeeding preference in this population (Galler, Harrison, Biggs, Ramsey, & Forde, 1999), delayed infant social, cognitive and motor development in the first 6 months of life (Galler et al., 2000), difficult infant temperament at 6 months (Galler, Harrison, Ramsey, Butler, & Forde, 2004) and lower scores on a national high school entrance examination at 11 years of age (Galler, Ramsey, et al., 2004). Maternal depression has been recognized in other studies as influencing maternal feeding attitudes (Tamminen, 1988) and also the duration of breastfeeding (e.g. Cooper, Murray, & Stein, 1993; Faleceto, Giugliano, & Fernandes, 2004). In the current study, we extend our earlier observations by describing feeding attitudes of mothers at 7 weeks postpartum and relating these views to maternal moods and feeding practices at 7 weeks, 3 months and 6 months postpartum. A prospective design was used to identify concurrent and predictive relationships over the first 6 months of life, when breastfeeding has been recognized as being especially important. The study addressed four major hypotheses: we predicted that (1) socioeconomic and demographic features would be closely associated with feeding attitudes; (2) feeding attitudes would be closely associated with maternal mood; (3) feeding attitudes would impact feeding practices in the first 6 months of life; and (4) mood and feeding attitudes would each contribute independently to breastfeeding. The early identification of factors associated with breastfeeding performance may serve as a basis for developing interventions aimed at increasing the number of mothers who breastfeed their infants in this setting.

3. Results 3.1. Relationships between Feeding Attitudes Scores and Background Variables The overall canonical correlation between feeding attitudes factors and demographic/home environmental factors was statistically significant (Rc = .40, p < .01). Protected multiple correlations between the group of feeding attitudes factors and the individual socioeconomic/home environment (SES/HE) factors confirmed significant associations for Family Income (Factor 3; R = .31, p < .001) and Information Seeking (Factor 9; R = .31, p < .01). Conversely, protected multiple correlations between 3 individual feeding attitudes, Breastfeeding is Attractive (Factor 2; R = .33, p < .01), Superstitions about Breastfeeding (Factor 5; R = .30, p < .05), and Breastfeeding is Better (Factor 6; R = .30, p < .05) and the group of SES/HE factors were also significant. Corresponding step-down Pearson correlations confirmed that mothers with higher incomes relative to those with lower incomes were more likely to have believed that breastfeeding was better for their infants and that breastfeeding made a woman more physically attractive. Women with higher incomes also had fewer superstitions about breastfeeding than did mothers with lower incomes. Additionally, mothers who sought out more information were more likely to have fewer superstitions about breastfeeding, but less likely to feel that breastfeeding was physically attractive. Feeding attitudes were also assessed in relationship to pregnancy history. The overall canonical correlation between the group of feeding attitudes factors and pregnancy history was significant (Rc = .30, p < .05). First, we correlated the individual pregnancy variables with the group of feeding attitudes factors. Protected multiple correlations confirmed that maternal age at first pregnancy (but not gestational age, birth order or parity) was the only pregnancy-related variable that was significantly correlated with the group of feeding attitudes factors (R = .27, p < .01). Conversely, protected multiple correlations between the individual feeding attitudes and group of pregnancy variables showed that Superstitions about Breastfeeding (Factor 5) and Breastfeeding is Better (Factor 6) were both significantly associated with the pregnancy variables (R = .24, p < .05 and R = .25, p < .05, respectively). The corresponding protected partial correlations indicate that the belief that breastfeeding was better for their infants correlated with maternal age at first pregnancy (r = .17, p < .05); and that superstitions about breastfeeding (Factor 5) was negatively correlated with maternal age at first pregnancy (r = −.15, p < .05). The feeding attitudes factors were also examined in relationship to infant size during the postpartum period. Feeding attitudes were not significantly associated with infant lengths or infant weights at birth, 7 weeks, 3 or 6 months. Maternal feeding attitudes also had no significant relationship with maternal anthropometry at 6 months. Finally, a multivariate F-test for the group of feeding attitude factors, and follow-up t-tests for each individual feeding attitude factor indicated that feeding attitudes of mothers of boys did not differ significantly from those expressed by mothers of girls. In summary, maternal attitudes about infant feeding were associated with Family Income and Information Seeking (SES/HE Factors 3 and 9) and maternal age at first pregnancy. All subsequent analyses described below were, therefore, performed both with and without correcting for these background variables. 3.2. Feeding attitudes associated with maternal moods Next, we assessed associations between maternal feeding attitudes at 7 weeks and maternal moods at 7 weeks and 6 months postpartum. Significant correlations were found between the feeding attitudes factors and maternal moods at both postpartum ages even after correcting for the background variables (7 weeks, Rcp = .44, p < .01 and 6 months, Rcp = .45, p < .001). Details of the 7 week analyses are presented in Table 3. The right hand column shows the protected multiple correlations between the group of feeding attitudes and each individual maternal mood (corrected and uncorrected for the background variables). As may be seen in the table, the group of feeding attitudes factors were significantly associated with three maternal moods, including Despair (p < .001), Zung Depression scores (p < .05) and Trust (p < .05), both before and after controlling for the background variables. The bottom row of Table 3 shows multiple correlations between the individual feeding attitudes and the group of six maternal mood scores. Two feeding attitudes factors, Breastfeeding is Better (Factor 6; p < .05) and Breastfeeding is Restrictive (Factor 4, p < .05), were each significantly associated with the group of maternal mood scores. Step-down Pearson correlations confirmed that mothers who reported feeling despair and depression at 7 weeks, as compared with women reporting fewer depressive symptoms at this time, were less likely to believe that breastfeeding was better for their infants. They were also more likely to be superstitious about breastfeeding (and, more specifically, to have no confidence that breast feeding leads to more mother–infant bonding). Mothers with more symptoms of depression and anxiety were also more likely to report that breastfeeding was restrictive (and specifically, that it made them look dowdy and that is restricted their freedom). Results of the corresponding 6-month analyses also confirmed a significant correlation between maternal Despair, Zung Depression and Anxiety scores and the group of maternal feeding attitudes factors, similar to the 7 weeks results (details not presented). Table 3. Feeding attitudes by maternal mood at 7 weeks, controlled for family income, information seeking, and age at first pregnancy (n = 145) Maternal mood factors Feeding attitudes factors Multiple R Breastfeeding should be Private (Factor 1) Breastfeeding is Attractive (Factor 2) Doctors Recommend Breastfeeding (Factor 3) Breastfeeding is Restrictive (Factor 4) Superstitions about Breastfeeding (Factor 5) Breastfeeding is Better (Factor 6) Morale Scale: Despair (Factor 1) – [–] – [–] – [–] – [–] .20*[.26**] −.28***[−.34***] .40***[.47***] Enjoys Children (Factor 2) – [–] – [–] – [–] – [–] .16†[.17*] – [–] .20 [.20] Trust (Factor 3) .16† [–] – [–] – [–] – [–] – [–] −.17*[−.17*] .30*[.29†] Self-Justification (Factor 4) – [–] – [–] – [–] −.20*[−.19*] – [–] – [–] .23 [.24] Zung Depression Scale .19*[.19*] – [–] – [–] .18*[.17*] – [.18*] – [–] .31*[.37**] Zung Anxiety Scale .17*[.17*] – [–] – [–] .17* [.15†] – [–] – [–] .26 [.30*] Multiple R .29† [.27†] .17 [.18] .12 [.09] .31 [.30*] .28†[.31*] .33*[.39***] Rcp = .44 **[Rc = .49 ***] Values in square parentheses are uncontrolled. (a) Boldfaced values are statistically significant. R indicates multiple correlations. Other values are simple Pearson correlations (r values) or canonical correlations (Rc). (b) Because of the low reliabilities of Factors 4, 5, and 6, we correlated their component items with mood and found that, for Factor 4, “Bottle-feeding makes it easier to have a career”, “Breastfeeding makes you look dowdy” and “Breastfeeding restricts your freedom” were correlated with depression. For Factor 5, “Doubts that breastfeeding creates a healthy mother–infant bond” correlated with despair. For Factor 6, “Bottle-feeding is just as good as breast feeding” correlated with depression, anxiety and despair. * p < .05. ** p < .01. *** p < .001. † .05 < p < 0.1. Table options Fig. 1 shows the inverse relationships between the perception that Breastfeeding is Better (Factor 6) at 7 weeks and maternal Despair at both 7 weeks and 6 months. Mothers who reported more symptoms of despair at both infant ages were also less likely to believe that breastfeeding was better for their infants at 7 weeks. Of note, this relationship was linear at both postpartum ages, showing associations across the range of scores even among women who did not have overt symptoms of clinical depression. Thus, maternal moods were clearly associated with their infant feeding attitudes. Maternal belief that breastfeeding is better for the infant at 7 weeks ... Fig. 1. Maternal belief that breastfeeding is better for the infant at 7 weeks associated with maternal moods at 7 weeks (r = −.34, p < .001) and 6 months (r = −.17, p < .05). (Number of subjects at 7 weeks: n1 = 22, n2 = 43, n3 = 61, n4 = 24; 6 months: n1 = 25, n2 = 53, n3 = 70, n4 = 27). Figure options 3.3. Feeding attitudes associated with feeding practices In order to determine the relationship between women's feeding attitudes and actual feeding practices, we compared feeding attitudes at 7 weeks postpartum with feeding practices assessed at 7 weeks, 3 and 6 months postpartum. As may be seen in Table 4, a significant overall canonical correlation was present between the group of feeding attitudes factors at 7 weeks and the group of feeding practices factors at 3 months both before and after correcting for the background variables (Rc = .48, p < .01; Rcp = .47, p < .01). There were no significant associations between feeding attitudes and practices measured at 7 weeks or at 6 months. The right-hand column of Table 4 shows multiple correlations between the group of feeding attitudes factors and individual feeding practices factors. As may be seen in the table, multiple correlations between the group of feeding attitudes and two feeding practices, Mother Breastfeeds Baby (Factor 2) and Baby Sleeps in Own Bed (Factor 4) were significant (p < .01 and p < .05, respectively, after controlling for background variables). The bottom row of the table shows multiple correlations between individual feeding attitudes and the group of feeding practices factors. As shown in the table, significant multiple correlations between Breastfeeding Should Be Private (Factor 1) and Breastfeeding is Better (Factor 6) and the group of feeding practices were present (both p < .01 after controlling for the background variables). Corresponding protected step-down Pearson correlations confirmed that mothers who had believed that breastfeeding was better for their babies at 7 weeks were significantly more likely to breastfeed. Further, that their infants were reported as having less intense suckling responses at 3 months. In addition, mothers who believed that breastfeeding should be private were less likely to breastfeed and more likely to sleep with their infants. Table 4. Maternal feeding attitudes by feeding practices at 3 months, controlled for family income, information seeking, and age at first pregnancy (n = 168) Feeding practices factors Feeding attitudes factors Multiple R Breastfeeding should be Private (Factor 1) Breastfeeding is Attractive (Factor 2) Doctors Recommend Breastfeeding (Factor 3) Breastfeeding is Restrictive (Factor 4) Superstitions about Breastfeeding (Factor 5) Breastfeeding is Better (Factor 6) Baby prefers breastfeeding (Factor 1) −.14† [−.14†] – [–] −.13∼ [–] – [–] – [–] – [.14†] .26† [.26†] Mother breastfeeds baby (Factor 2) −.16*[−.15*] – [–] – [–] – [–] – [–] .25**[.22**] .33**[.30*] Baby's feeding intensity (Factor 3) – [–] – [–] – [–] – [–] – [–] −.20**[−.22**] .23 [.25] Baby sleeps in own bed (Factor 4) −.19*[−.17*] – [−.16*] – [–] – [–] –[–] – [.18*] .28*[.31*] Relatives feed baby (Factor 5) – [–] – [–] – [–] – [–] – [–] – [–] .15 [.16] Multiple R .31**[.29*] .17 [.19] .16 [.16] .11 [.13] .19 [.14] .32**[.36***] Rcp = .47 **[Rc = .48 **] Values in square parentheses are uncontrolled. (a) Boldfaced values are statistically significant. R indicates multiple correlations. Other values are simple Pearson correlations (r values) or canonical correlations (Rc). (b) Because of the low reliability of Factor 6, we correlated its component items with the feeding practices factors and found that one item, “Bottle-feeding is just as good as breast feeding,” correlated best with the feeding practices. * p < .05. ** p < .01. *** p < .001 † .05 < p < 0.1. Table options Fig. 2 illustrates the concurrent and predictive relationships between maternal beliefs that Breastfeeding Is Better for the infant (Factor 6) at 7 weeks and the extent to which Mothers Breastfed Their Babies (Factor 2) at 7 weeks, 3 and 6 months. As may be seen, there was a linear relationship between the belief that Breastfeeding is Better and the actual amount of breastfeeding across the range of factor scores at all three postpartum ages. The multiple correlation between Breastfeeding is Better (Factor 6) and the amount of breastfeeding across the three ages was significant (Rp = .27, p < .05). Individual Pearson correlations were all significant (p < .05) relative to amounts of breastfeeding at 7 weeks, 3 and also 6 months (when only 34% of women were still breastfeeding). Thus, the belief that breastfeeding is better for the infant at 7 weeks postpartum predicted concurrent and subsequent feeding practices up to 6 months. Mother's belief that breastfeeding is better at 7 weeks predicts amount of ... Fig. 2. Mother's belief that breastfeeding is better at 7 weeks predicts amount of breastfeeding at 7 weeks (rp = .20, p < .05), 3 months (rp = .26, p < .01), and 6 months (rp = .18, p < .05) (N = 113; number of subjects at each standard deviation group: n1 = 18, n2 = 32, n3 = 45, n4 = 18). Figure options 3.4. Independent contributions of feeding attitudes and maternal mood to breastfeeding Finally, because of the significant associations between maternal mood and feeding attitudes and also maternal mood and the baby's preference for breastfeeding (Galler et al., 1999), it was possible that maternal mood mediated the relationship between feeding attitudes and feeding practices. In order to determine whether maternal symptoms of despair accounted for the relationship between feeding attitudes at 7 weeks and the amount of breastfeeding, we re-ran the above analyses controlling for Maternal Despair scores at 7 weeks and the other background variables reported as having an association with feeding attitudes. Even after correcting for these variables, the group of feeding attitudes at 7 weeks continued to have a significant association with the feeding practices at 3 months (Rcp = .54; p < .05), similar to the uncorrected analyses above. Again, the amount of breastfeeding (Feeding Practices Factor 2) was correlated with the group of feeding attitudes, especially the belief that Breastfeeding is Better for the infant (Rp = .34; p < .05). Similar findings were also obtained when we corrected for the Zung Depression scores. These findings confirm that mothers who had believed breastfeeding was better and who felt that breastfeeding was not restrictive were also more likely to breastfeed, even after maternal mood was taken into account. In our prior study using a shorter version of the feeding questionnaire, we reported that maternal mood impacted the infant's preference for breastfeeding (Galler et al., 1999). In order to exclude the possibility that feeding attitudes mediated this relationship, we re-examined the relationship between maternal moods and feeding practices, controlling for Breastfeeding Should Be Private (Factor 1) and Breastfeeding is Better (Factor 6). Prior to controlling for the feeding attitudes factors, significant associations were present between maternal moods at 7 weeks and feeding practices at 7 weeks (Rcp = .46; p < .05) and also at 3 and 6 months (Rcp = .53 and Rcp = .39, both p < .01). However, after correcting for the two feeding attitudes factors (and all other background variables), the group of maternal moods at 7 weeks was correlated with the group of Feeding Practices Factors only at 6 months (Rcp = 0.42; p < .01), with 3 months being a trend (Rcp = 0.52; 1. > p > .05). Significant multiple correlations between the group of 7 week maternal moods and Baby Prefers Breastfeeding (Factor 1) were present at both 3 and 6 months (Rp = .38 and .33, both p < .05), but not with Mother Breastfeeds Baby (Factor 2). These findings show that, while maternal feeding attitudes and moods were closely associated at 7 weeks, at 3 months and at 6 months, feeding attitudes and moods contributed independently to infant feeding practices and, in fact, contributed to different aspects of infant feeding especially at 6 months.