![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
,**
* Departments of Anthropology,
International Health and
** Epidemiology, Emory University, Atlanta, GA 30322
2To whom correspondence should be addressed. E-mail: dsellen{at}emory.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT SUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
KEY WORDS: • full breast-feeding • lactogenesis • onset of lactation • predominant breast-feeding • survival analysis.
Despite recommendations to provide infants with breast milk exclusively for the first 6 mo (1), nonbreast milk substances are often introduced within weeks after birth (2,3). Nonbreast milk substances are more likely than breast milk to carry infections to the infant, and decreasing breast milk uptake may reduce breast milk’s immunological benefits to the infant and its contraceptive effect to the mother (4–6). Considering these potentially deleterious effects of early supplementation with nonbreast milk substances, it is important to understand what factors lead to their early introduction.
Several categories of breast-feeding practice are defined by the WHO (7). Exclusive breast-feeding indicates that the infant has received only breast milk and no other liquids or solids with the exception of drops or syrups consisting of vitamins, mineral supplements or medicines. Predominant breast-feeding describes a pattern in which breast milk remains the predominant source of nourishment, but one in which water, water-based drinks (e.g., sweetened and flavored water, teas or infusions), fruit juice, oral rehydration salts solution, drops and syrup forms of vitamins, minerals and medicines, and limited quantities of ritual fluids are fed. A third category, full breast-feeding indicates that an infant is either exclusively or predominantly breast-fed. In many regions around the world, as is the case in this sample, exclusive breast-feeding is so rare by the end of mo 1 of life (2,3) that maintenance of full breast-feeding may serve as a more useful category for distinguishing relatively healthy and unhealthy feeding practices during the first 6 mo of life (8–10). Although little research has examined the effect of full breast-feeding (vs. exclusive or any breast-feeding) on infant health, most of the immunological, nutritional, and maternal contraceptive benefits of exclusive breast-feeding are likely to extend to infants who are fully fed breast milk (8–10).
Onset of lactation (OL) has been defined as the "initiation of copious milk production in the mammary gland" (11) and measured as the time at which women report a perception that their breast milk has "come in," based on cues such as breast hardness, fullness/heaviness, or swelling and leakage of colostrum or breast milk (12,13). Maternal perception of "milk coming in" is itself a valid clinical indicator (13,14) of lactogenesis stage II, the secretion of mature milk marked by changes in salt, sugar and protein composition that occur 32–40 h postpartum (pp) (15,16).
The timing of OL has been shown to be related to breast-feeding outcomes, such as perceived insufficient milk (17), the preonset introduction of breast milk supplements (18), and the timing of supplementation several months after delivery (13,19–21). Several mechanisms have been proposed to explain the statistical associations observed between delayed OL and the timing of later supplementation (Fig. 1). First, mothers who lack an understanding of the stages of milk production and who experience later OL may lose confidence in their ability to produce adequate amounts of breast milk. Consequently, they may supplement their infant believing that the infant cannot survive on the small quantity of colostrum secreted during the first days after delivery. If this loss of confidence continues beyond OL, it may also increase the risk of early supplementation with formula or cereal-based fluids at later ages. Furthermore, deciding to supplement in the first days after delivery because of delayed OL may prime future decisions to supplement, leading to an earlier introduction of formulas or cereal-based fluids (18,20,22). Second, the introduction of supplements in the first days after delivery may independently delay lactogenesis stage II and accelerate cessation of breast-feeding. This hypothesis draws on the idea that milk removal during early infant suckling may elicit physiologic responses that trigger earlier OL while independently creating a stable pattern of suckling and milk production that leads to successful long-term breast-feeding (12,13,15,23). Introduction of nonbreast milk substances may reduce the frequency and duration of suckling, thus independently disrupting the process leading to OL and serving as the initial step toward more frequent supplementation and greater breast milk displacement in the first 6 mo. Third, Perez Escamilla et al. (18) proposed that mothers who are biologically predisposed to experience delayed OL may also be more likely to introduce supplements and to end breast-feeding early.
|
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Data for this analysis were collected as part of a longitudinal, multidisciplinary nutrition intervention study begun in the mid-1960s (24–26). The study focused on four predominantly Ladino (i.e., mestizo) communities in rural eastern Guatemala centered around villages with long-running administrative identities (27–29).
Breast-feeding data were collected prospectively for 501 infants born between June 1, 1996 and June 30, 1999 (84.2% of all recorded births). Trained interviewers (who were rotated among the communities) administered retrospective questionnaires every 2 wk to caregivers in their homes. For each of the previous 14 d, feeding status was assigned to one of five categories (no milk-feeding; breast-feeding with liquids; breast-feeding with formula; breast-feeding or formula with solids; and exclusive breast-feeding).
Additional information on maternal, infant and birth variables was collected for the entire sample during a census conducted in 1996 and during prenatal visits. Detailed data on feeding in wk 1 were collected by maternal recall at 8 d postpartum for a random subsample of infants (n = 338). After exclusion of nine cases with incomplete data and one case in which breast-feeding was never initiated, the analysis considered 328 infants. Institutional Review Boards at the Instituto de Nutrición de Centro América y Panamá (INCAP) and the Emory University School of Medicine approved the research protocol.
Dependent variable.
We categorized infants as fully breast-fed when breast milk was reportedly the predominant source of nourishment. This category allowed supplementation with limited quantities of water-based drinks, fruit juice, oral rehydration salts, ritual fluids and drops of medicine, but excluded the introduction of other nutritive substances such as nonhuman milk, cereal-based fluids, semisolids and solids (7,30). Interviews every 2 wk with mothers recorded daily information on infant feeding status; thus, the date at which breast-feeding practices changed could be determined with high precision. We retained for this analysis only data between birth and 6 mo because this is the period in which exclusive or full breast-feeding is recommended for most infants (1).
The time at cessation of full breast-feeding was defined as the day that formula and/or cereal-based fluids were first introduced for a period lasting at least 3 d. This definition allowed mothers to deviate briefly (1 to 2 d) from full breast-feeding without being considered as having ended full breast-feeding. The distribution of the outcome variable derived with this definition was not different from those of other similarly derived variables (such as "the day that formula and/or cereal-based fluids are first introduced for a period lasting at least 7 d").
Cases in which full breast-feeding continued to the end of participation in the study were censored in survival analysis (n = 115). A majority of infants (n = 73) were censored because they had continued full breast-feeding to 6 mo. Twelve cases were lost to follow-up (2 infant deaths; 10 for unknown reasons). Also censored were cases in which mothers delivered <6 mo before the study end and continued full breast-feeding to the last visit (n = 30).
Independent variables.
Mothers were asked at 7 d pp what day they perceived OL (Spanish: ¿que dia le bajó la leche?) using a scale from 1 to 7 with 1 indicating the 24-h period beginning at midnight of the day the infant was born. The OL measure followed a question about the time of the infant’s birth. Timing of onset was coded dichotomously using a conventional cutoff (OL 1–3 d pp vs. OL > 3 d pp) used in recent studies of OL and breast-feeding (12,13,18,20).
In addition to the main predictor variable, 15 variables identified from earlier studies as potential confounders or effect modifiers were included in the statistical models. These included maternal [age at birth of infant, socioeconomic status (SES), education, workplace, marital status, parity, BMI at first pregnancy visit, and community of residence], infant (sex, birth weight) and birth variables (Cesarean section, place of birth, and preterm birth, i.e., gestational age < 260 d). Gestational age was estimated from the reported date of last menstruation (LMP) for a large majority (>93%) of cases. Gestational age was estimated from ultrasound in the second trimester for a small minority of mothers for whom LMP was not available or was largely discrepant with the estimate from ultrasound. To assess the effect of supplementation that preceded OL, "preonset supplementation" (i.e., not necessarily prelacteal feeding) was coded as 1 if an infant was supplemented with nonbreast milk substances before maternal OL, or 0 otherwise.
Statistical analysis.
SAS version 8.0 statistical software (31) was used for all statistical analyses. To minimize potential bias due to systematic missing data and to include as many cases as possible in the analysis, a category of "missing" was created for each covariate of interest. For "education" this resulted in a 5-level categorical variable (<3rd grade, 3rd–6th grade, finished 6th grade, >6th grade, missing). Under the assumption that "missingness" is random with respect to the outcome variable, one would expect the "missing" category to have an effect size within the range of those observed for the nonmissing stratum (32).
To examine the relationship between delayed OL and the timing of full breast-feeding cessation, Cox proportional hazards regression models were fit with SAS 8 PROC PHREG (33). Cox regression allows inclusion of censored cases in statistical models and adjustment of estimates for potential confounders and effect modifiers (33). Each covariate was examined individually and in a multiple-adjusted model to determine whether there was a violation of the proportional hazards assumption (using a goodness-of-fit test at P < 0.05). Covariates that did not satisfy the proportional hazards assumption were included in the model as strata. Potential interaction was assessed for individual multiplicative interaction terms (e.g., community x delayed OL) by comparing the log likelihood statistics of models with and without the interaction terms. Interaction terms were retained in the final model at the P < 0.10 level.
Daily feeding data from wk 1 after birth allowed us to assess whether early supplementation delays OL or whether delayed OL leads to supplementation. For example, we could estimate how strongly supplementation at d 1 predicted OL at d 2, by assessing the relative odds of OL at d 2 for mothers who did and did not supplement at d 1. Conversely, we could determine whether not perceiving OL on a given day increased the odds of supplementation on the following day by assessing the relative odds of beginning supplementation at d 3 for infants whose mothers did or did not perceive OL on d 1 or 2.
To explore whether estimates from all analyses were biased due to either double-counting or correlation within sibling sets, we performed the same analyses on randomly chosen subsamples with only one sibling from each set.
|
RESULTS |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
30 kg/m2). Home births were common (42.4%) and nearly twice as likely in community A than in the other communities. Among infants with birth weight information, 10.3% were recorded as low birth weight (<2500 g). The proportion of Cesarean sections was 10.7% and that of preterm births was 7.8%. There were 38 sibling pairs and one set of three siblings in the final sample. As is commonly found in similar studies, there was no significant correlation (r = 0.007, P > 0.20) between full breast-feeding duration for younger and older sibs within nontwin noncensored sibling pairs (n = 25), and results from random subsamples with only one sibling per subsample did not differ from results on the full sample. We therefore present results retaining all eligible children and include siblings.
|
Only one mother did not initiate breast-feeding, and she was excluded from analysis. Over half (52.4%) of infants were first put to the breast within h 1 of life. A minority (10.1%) perceived OL later than 3 d pp. A substantial proportion of infants (30.1%) were fed nonbreast milk substances (other than medicines or oils) before the OL. Sugar water was the most common preonset supplement (22.3%). Preonset feeding with water (4.0%), nonnutritive liquids and teas (4.5%) and formula (2.1%) was also reported. Note that feeding of these substances (except formula) led to categorization of an infant as fully rather than exclusively breast-fed. The proportion of mothers supplementing before OL varied significantly across communities.
Full breast-feeding.
Among the 328 cases retained for the analysis, missing data in daily records before full breast-feeding accounted for 12.1% of total days sampled (4886/40508). Although most mothers ended exclusive breast-feeding within mo 1 of life or never began, most also continued any breast-feeding past the first 6 mo of life (Fig. 2). Median duration of full breast-feeding was 5 mo, and a minority (25.5%) of infants who were not censored before 6 mo continued full breast-feeding to 6 mo or longer.
|
A number of covariates (maternal age and education level, Cesarean section, birth at term, preonset supplementation and delayed OL) were significantly and independently associated with an increased hazard of ending full breast-feeding during the first 6 mo (Table 2). Community membership was the only covariate for which there was a significant interaction with delayed OL (-2 log likelihood difference = 15.863, df = 3, P = 0.001, 13 interactions tested). The interaction is represented in Table 2 with community-specific hazards ratios.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
These data do not support the hypothesis (hypothesis 2, Fig. 1) that preonset supplementation leads to the later onset of lactation and to a shorter duration of full breast-feeding. Rather, they provide evidence that later onset of lactation leads to supplementation in wk 1 and can lead to earlier cessation of full breast-feeding (hypothesis 1, Fig. 1). Further evidence suggests that the influence of OL on full breast-feeding cessation is mediated in part by preonset supplementation (hypothesis 1, Fig. 1).
The observation that delayed onset of milk puts mothers at greater risk of perceived milk insufficiency and losing confidence in their capacity to breast-feed their infant (17,37,38) is consistent with hypothesis 1. Indeed, one third of mothers in this study who supplemented in wk 1 reported insufficient milk as a reason for supplementation (39). Thus, rather than suggesting that early suckling and breast milk removal separately influence two dimensions of breast-feeding practice (early perceived onset and late full breast-feeding cessation), these data suggest that a single psychophysiologic phenomenon (timing of perceived onset) influences both timing of first supplementation in the short term and the hazard of ending full breast-feeding in the long term.
The observation that younger mothers or mothers with more years of education ended full breast-feeding earlier during the first 6 mo of their infant’s life is consistent with other research in developing regions (40). Preterm birth had a protective effect on full breast-feeding, possibly because mothers viewed premature babies as having special needs for frequent breast-feeding or as being more vulnerable to supplementation with nonbreast milk substances. An alternative explanation is that small or slow-growing babies were not introduced to nonbreast milk foods as early because their mothers were less concerned about producing enough to keep up with the fast growth or the needs of a large child. As consistently found wherever studied, Cesarean birth put an infant at significantly greater hazard of ending full breast-feeding during the first 6 mo (41).
The associations between adequate early feeding and the maintenance of full breast-feeding appeared to be modified by unmeasured local factors specific to different communities. The HR were positive and significant in two communities, null in one and inverse in the fourth. This last HR was not significant and is biologically implausible. Although the community interaction term was significant (P < 0.001), the four community-specific estimates of the HR were imprecise, due to small sample sizes (52–102 women per community, of whom 8 to 9 women per community experienced delayed OL) and the result could be due to chance alone. The study was not designed to identify the specific causes of any community-level differences found. However, a variety of factors related to family income, maternal work organization, sanitation and economic and social integration (all of which likely varied across the communities) may influence both early and later breast-feeding practices (27–29). For example, relative to the other two communities in the study, the two communities in which delayed OL appeared to undermine the persistence of full breast-feeding were more economically developed (27–29), with higher mean household income, more diverse job opportunities (in nearby factories and mechanized agriculture, and through more frequent and direct bus connections to Guatemala City) and a larger population (>1500). They also had higher rates of births at home. These and other characteristics may combine to produce different local contextual factors influencing maternal intention to breast-feed, experience of timely OL, response to delayed OL and support for full breast-feeding. For example, mothers in the more economically developed communities may have had access to a wider range of alternative substances used to supplement an infant if perception of late onset undermines long-term confidence in the ability to breast-feed. Alternatively, new mothers in communities with a more traditional economy may be more likely to have jobs that are compatible with breast-feeding, to be supported by local breast-feeding advocates or to be less influenced by negative images of breast-feeding. These and other plausible explanations should be explored in future studies.
The present study considered breast-feeding data that are unique in non-Western contexts for the combination of survey frequency (every 2 wk) and sample size (n > 300). Few studies have followed day-to-day variation in breast-feeding practices over an extended period of time, and most of these have occurred in Western settings (42). With these data, it was possible to determine how long a specific pattern of breast-feeding practice continued without interruption, rather than whether it was present at a certain time point during the study (36); it avoids the effects of "heaping" characteristic of retrospective studies involving longer term recall (43). Finally, as part of a larger intergenerational study, the breast-feeding data were linked to a large number of maternal-, birth- and infant-related covariates, which were used to assess a wide range of confounding influences.
OL has been validated as a proxy for onset of lactogenesis II (13,14), with the Spanish term "me bajó la leche" a culturally valid signifier of the event in these Guatemalan communities. Less certain is the manner in which the variable "full breast-feeding" is used. The WHO definition of full breast-feeding requires that breast milk be the predominant source of nutrition for an infant, but data from the surveys every 2 wk provided information only on what types of substance were fed rather than relative proportions (7,10). In addition, the definition of full breast-feeding cessation used in this analysis allowed only a 2-d violation of full breast-feeding before full breast-feeding was assumed to have ended. In several instances, infants who ended full breast-feeding according to this definition did return to full breast-feeding for extended periods after being recorded as having switched to partial or no breast-feeding (44). Further studies are required to determine the most effective decision rule for defining the cessation of full breast-feeding.
Both the predictor (timing of OL) and main mediating variable (preonset supplementation) were assessed using a recall questionnaire administered 7 d after birth. Recall bias may therefore have been introduced in the estimation of these variables. Specifically, mothers who were more pessimistic about their ability to exclusively breast-feed the infant in wk 1 may have recalled earlier supplementation and later OL than their more optimistic coparticipants. This recall bias may also have influenced inferences about the temporal priority of timing of OL and supplementation. However, Pérez-Escamilla and Chapman (45) showed that women in Connecticut can recall timing of OL 6 mo after delivery with a very high degree of sensitivity and a reasonable level of specificity, suggesting that OL recalled 
7 d after delivery may not be associated with a high degree of misclassification when categorized dichotomously as 
3 d pp and >3 d pp. The assessment of OL timing by days (rather than hours) after birth creates another potential source of misclassification with OL on d 1 potentially covering 0–24 h pp, OL on d 2 potentially covering 0–48 h pp, OL on d 3 potentially covering 24–72 h pp. As a source of random error, this form of misclassification should generate a bias towards the null.
These findings provide new and detailed evidence that the pattern of feeding infants in the first few days of life is a critical determinant of subsequent feeding practices. Specifically, delayed OL puts infants at significantly greater hazard of ending full breast-feeding during the first 6 mo. Importantly, the early use of even small quantities of water-based fluids, fruit juices and ritual fluids may have adverse effects for infants of mothers who experience delayed OL. Further research should aim to discover why some mothers respond to delayed OL by initiating a cascade of nonbreast milk supplementation, why some mothers experience delayed OL in the first place and why the relationship between delayed onset of lactation and full breast-feeding holds only in certain communities. Indeed, studying the communities in which delayed OL does not lead to an increased hazard of ending full breast-feeding may reveal protective practices that could be extended to communities in which delayed OL increases the hazard of ending full breast-feeding during the first 6 mo. The results of such further research would inform efforts to develop locally appropriate breast-feeding support for those mothers who do experience late onset of lactation.
|
FOOTNOTES |
|---|
3 Abbreviations used: HR, hazards ratio; LMP, last menstrual period; OL, onset of lactation; OR, odds ratio; pp, postpartum; SES, socioeconomic status.
4 Percentages include only those cases with nonmissing values in the denominator.
Manuscript received 16 April 2003. Initial review completed 4 May 2003. Revision accepted 4 June 2003.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
1. World Health Organization (2001) The Optimal Duration of Exclusive Breastfeeding 2001 WHO Geneva, Switzerland.
2. World Health Organization (1996) Global Data Bank on Breastfeeding 1996 Nutrition Unit WHO, Geneva, Switzerland.
3. Haggerty, P. A. & Rutstein, S. O. (1999) Breastfeeding and Complementary Infant Feeding, and the Postpartum Effects of Breastfeeding 1999 Macro International Calverton, MD.
4. Feachem, R. & Koblinsky, M. (1983) Interventions for the control of Diarrheal Diseases among young children. Bull. WHO 61:641-652.[Medline]
5. Onnela, T. (1997) Persistent Diarrhea and Breastfeeding 1997 World Health Organization Geneva, Switzerland.
6. Goldman, A., Garza, C., Nichols, B. & Goldblum, R. (1982) Immunological factors in human-milk during the 1st year of lactation. J. Pediatr. 100:563-567.[Medline]
7. World Health Organization (1991) Indicators for Assessing Breast-Feeding Practices 1991 Division of Diarrhoeal and Acute Respiratory Disease Control. WHO Geneva, Switzerland.
8. Padmadas, S. S., Hutter, I. & Willekens, F. (2002) Weaning initiation patterns and subsequent linear growth progression among children aged 2–4 years in India. Int. J. Epidemiol. 31:855-863.
9. Hop, L. T., Gross, R., Giay, T., Sastroamidjojo, S., Schultink, W. & Lang, N. T. (2000) Premature complementary feeding is associated with poorer growth of Vietnamese Children. J. Nutr. 130:2683-2690.
10. Eckhardt, C., Rivera, J. A., Adair, L. S. & Martorell, R. (2001) Full breast-feeding for at least four months has differential effects on growth before and after six months of age among children in a Mexican community. J. Nutr. 131:2304-2309.
11. Grajeda, R. & Perez-Escamilla, R. (2002) Stress during labor and delivery is associated with delayed onset of lactation among urban Guatemalan women. J. Nutr. 132:3055-3060.
12. Chapman, D. J. & Perez-Escamilla, R. (1999) Identification of risk factors for delayed onset of lactation. J. Am. Diet. Assoc. 99:450-454.[Medline]
13. Chapman, D. J. & Perez-Escamilla, R. (2000) Maternal perception of onset of lactation is a valid, public health indicator of lactogenesis stage II. J. Nutr. 130:2972-2980.
14. Perez-Escamilla, R. & Chapman, D. J. (2001) Validity and public health implications of maternal perception of the onset of lactation: an international analytical overview. J. Nutr. 131:3021S-3024S.
15. Neville, M. C. & Morton, J. (2001) Physiology and endocrine changes underlying human lactogenesis II. J. Nutr. 131:3005S-3008S.
16. Lawrence, R. A. (1994) Physiology of human lactation. Breastfeeding: A Guide for the Medical Profession 1994:59-89 Mosby St. Louis, MO.
17. Segura-Millan, S., Dewey, K. G. & Perez-Escamilla, R. (1994) Factors associated with perceived insufficient milk in a low-income urban population in Mexico. J. Nutr. 124:202-212.
18. Perez-Escamilla, R., Segura-Millan, S., Canahuati, J. & Allen, H. (1996) Prelacteal feedings are negatively associated with breast-feeding outcomes in Honduras. J. Nutr. 126:2765-2773.
19. Perez-Escamilla, R., Segura-Millan, S., Pollitt, E. & Dewey, K. G. (1993) Determinants of lactation performance across time in an urban population from Mexico. Soc. Sci. Med. 37:1069-1078.
20. Chapman, D. J. & Perez-Escamilla, R. (1999) Does delayed perception of the onset of lactation shorten breastfeeding duration?. J. Hum. Lact. 15:107-111.
21. McCabe, T. L. (1982) The Assessment of Factors That Affect the Duration of Lactation of Women Residing in Eastern Connecticut: A Predictive Model 1982 Doctoral thesis University of Connecticut, Storrs, CT.
22. Sellen, D. W., Thompson, A. L., Hruschka, D. J., Stein, A. D. & Martorell, R. (2003) Early determinants of non-exclusive breastfeeding among Guatemalan infants. Pickering, L. K. Morrow, A. L. Schanler, R. J. Ruiz-Palacios, G. M. eds. Protecting Infants through Human Milk: Advancing the Scientific Evidence Base 2003 Kluwer Academic/Plenum Publishers New York, NY (in press). .
23. Neville, M. C., Morton, J. & Umemura, S. (2001) Lactogenesis: the transition from pregnancy to lactation. Pediatr. Clin. N. Am. 48:35-52.[Medline]
24. Read, M. S. & Habicht, J.-P. (1993) History of the INCAP longitudinal study on the effects of early nutrition supplementation in child growth and development. Food Nutr. Bull. 14:169-175.
25. Martorell, R., Habicht, J.-P. & Rivera, J. A. (1995) History and design of the INCAP longitudinal study (1969–77) and its follow-up (1988–89). J. Nutr. 125:1027S-1041S.
26. Habicht, J.-P. & Martorell, R. (1993) Objectives, research design, and implementation of the INCAP longitudinal field study. Food Nutr. Bull. 14:176-190.
27. Bergeron, G. (1993) Social and economic development in four Ladino communities of eastern Guatemala: a comparative description. Food Nutr. Bull. 14:221-236.
28. Engle, P. L., Carmichael, S. L., Gorman, K. & Pollitt, E. (1993) Demographic and socio-economic changes in families in four Guatemalan villages, 1967–1987. Food Nutr. Bull. 14:237-245.
29. Saenz de Tejada, S., Mazariegos, L. M., Asturias de Barrios, L. & Lara, J. L. (2002) Changes in the Socioeconomic and Cultural Conditions That Affect the Formation of Human Capital and Economic Productivity in Four Villages in Eastern Guatemala 2002 Instituto de Nutricion de Centro America y Panama Guatemala.
30. Cattaneo, A., Davanzo, R. & Ronfani, L. (2000) Are data on the prevalence and duration of breastfeeding reliable?. The case of Italy. Acta Paediatr. 89:88-93.
31. SAS Institute Inc (2000) SAS Procedures Guide, Version 8 2000 SAS Institute Inc. Cary, NC.
32. Thompson, W. D. & Kelsey, J. L. (1986) Methods in Observational Epidemiology 1986 Oxford University Press New York, NY.
33. Kleinbaum, D. G. (1996) Survival Analysis: A Self-Learning Text 1996 Springer Verlag New York, NY.
34. Perez-Escamilla, R., Cobas, J., Balcazar, H. & Holland Benin, M. (1999) Specifying the antecedents of breast-feeding duration in Peru through a structural equation model. Public Health Nutr 2:461-467.[Medline]
35. Jones, D., West, R. & Newcombe, R. (1986) Maternal characteristics associated with duration of breast-feeding. Midwifery 2:141-146.[Medline]
36. Aarts, C. (2000) How exclusive is exclusive breastfeeding? A comparison of data since birth with current status data. Int. J. Epidemiol. 29:1041-1046.
37. Greiner, T., Van Esterik, P. & Latham, M. C. (1981) The insufficient milk syndrome: an alternative explanation. Med. Anthropol. 5:233-260.
38. Tully, J. & Dewey, K. G. (1985) Private fears, global loss: a cross-cultural study of the insufficient milk syndrome. Med. Anthropol. 9:225-243.[Medline]
39. Sellen, D., Hruschka, D., Thompson, A., Neufeld, L. & Martorell, R. (2001) Perinatal Factors Associated with Deviation from Recommended Feeding Practices in Guatemalan Infants 2001 American Public Health Association Meetings Atlanta, GA.
40. Kocturk, T. & Zetterstrom, D. (1999) Thoughts about the rates of breastfeeding. Acta Paediatr 88:356-358.[Medline]
41. Chapman, D. J., Young, S., Ferris, A. M. & Perez-Escamilla, R. (2001) Impact of breast pumping on lactogenesis stage II after caesarean delivery: a randomized clinical trial. Pediatrics 107:E94.
42. Hornell, A., Aarts, C., Kylberg, E., Hofvander, Y. & Gebre-Medhin, M. (1999) Breastfeeding patterns in exclusively breastfed infants: a longitudinal prospective study in Uppsala, Sweden. Acta Paediatr 88:203-211.[Medline]
43. Perez-Escamilla, R., Lutter, C., Segall, A. M., Rivera, A., Trevino-Siller, S. & Sanghvi, T. (1995) Exclusive breast-feeding is associated with attitudinal, socio-economic and biocultural determinants in three Latin American countries. J. Nutr. 125:2972-2984.
44. Zohoori, N., Popkin, B. M. & Fernandez, M. (1993) Breastfeeding patterns in the Philippines: a prospective analysis. J. Biosoc. Sci. 25:127-138.[Medline]
45. Perez-Escamilla, R. & Chapman, D. J. (2001) Can women remember when their milk came in?. Adv. Exp. Med. Biol. 501:567-572.[Medline]
This article has been cited by other articles:
|
|
 |
|
  M. Chisenga, L. Kasonka, M. Makasa, M. Sinkala, C. Chintu, C. Kaseba, F. Kasolo, A. Tomkins, S. Murray, and S. Filteau Factors Affecting the Duration of Exclusive Breastfeeding Among HIV-Infected and -Uninfected Women in Lusaka, Zambia J Hum Lact, August 1, 2005; 21(3): 266 - 275. [Abstract] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
