Breast-Feeding
Intentions
and Practice
Among
Hispanic
Mothers
in Southern
California
Eunice Romero-Gwynn,
PhD, and Lucia Carias, BSFrom the Department of Nutrition, University of California, Davis
ABSTRACT. Breast-feeding intentions, breast-feeding in the hospital, and breast-feeding at home were studied among 132 Hispanic mothers participating in the Ex-panded Food and Nutrition Education Program in south-em California. There was not a large difference between total breast-feeding intention (77.7%) and total breast-feeding practice (63.8%). However, the 67.7% intention of exclusive breast-feeding drastically decreased to 19.7% and 17.2% in the hospital and at home, respectively. Formula supplementation increased by 4.5 times from intention to practice. Exclusive formula feeding increased from 10.0% to approximately 37.0% in the hospital and at home. Stepwise logistic regression identified that the likelihood of intending breast-feeding was greater for
mothers who migrated from Mexico than for mothers
born in the United States (odds ratio 4.75). The likelihood of breast-feeding practice was greater for mothers who initiated breast-feeding within the first 10 hours after birth as opposed to 11 or more hours (odds ratio 1.27),
for mothers who had a vaginal rather than cesarean
delivery (odds ratio 12.76), for mothers who did not return to work postpartum as opposed to working mothers (odds ratio 28.26), and for mothers who migrated from Mexico compared with mothers born in the United States (odds ratio 8.54). The importance of assessing and supporting mothers’ breast-feeding intentions in the pre- and post-partum period is documented. Training in the clinical aspects of breast-feeding and improvement of hospital protocols is recommended. Mothers intending to breast-feed should be identified and supported. Pediatrics
1989;84:626-632; breast-feeding, Hispanic women,
mi-grant, breast-feeding.
In the United States, a resurgence of
breast-feeding has occurred in the last 15 years,’ reversing
a downward trend initiated after World War II
which reached a low breast-feeding incidence of
approximately 25.0% in the early 19705.2 National
data from Martinez and Krieger1 for 1984 (collected
through mailed questionnaires) show an incidence
Received for publication Aug 15, 1988; accepted Oct 28, 1988. Reprint requests to (E.R.-G.) Dept of Nutrition, University of California, Davis, CA 95616.
PEDIATRICS (ISSN 0031 4005). Copyright © 1989 by the
American Academy of Pediatrics.
of in-hospital breast-feeding of 65.0% and 33.3%
among white and black mothers, respectively.
How-ever, no figures for Hispanics are provided in this
study.
Data from small studies of Hispanics show a large
variation in breast-feeding practice between
differ-ent Hispanic subgroups studied (ie, Mexicans,
Cu-bans, Puerto Ricans). Breast-feeding incidence for
women of Mexican descent living in the United
States has been reported by several authors.3
Fig-ures for breast-feeding intention vary from 22.6%
found by Rassin et a!4 in Texas to 82.0% reported
by Scrimshaw et al7 in Los Angeles. Data
concern-ing breast-feeding practice show variation from
31.1% to 60.0%. Smith et al3 found that 31.1% of
the Hispanic women studied between 1976 and 1979 along the United States/Mexican border breast-fed their infants. Samuels et al6 in Oakland, California,
and Young and Kaufman8 in North Carolina both
found a breast-feeding incidence of 60.0%.
Data concerning breast-feeding for Puerto
Ri-cans and Cubans is more limited. Figures reported
for Puerto Ricans by Mohrer9 in Connecticut and
by Bryant’#{176} in Florida are 11.0% and 10.0%,
re-spectively. The incidence for Cubans reported by
Bryant’#{176}in Florida is 12.0%. The extent to which variations in breast-feeding intention and practice
of Hispanics reported in the literature are
influ-enced by differences in definitions of terms and
research methodology is not clear.
Factors associated with breast-feeding have been
studied extensively. For black and white mothers,
the highest incidence of breast-feeding has been
found among college-educated women older than
20 years of age, with incomes of more than $15 000
per year. The most common factors associated with
breast-feeding among Hispanics are education,3’4”
mothers employment,7 infant birth order,4
grand-mother and members of network system,’#{176}
experi-ence with breast-feeding,3 and attitude of fathers’#{176}
and mothers toward breast-feeding.’2
The study reported in this paper provides
practices and variables related to these among a
group of Hispanic mothers (Mexican and Mexican
descent) living in three counties in southern
Cali-fornia. The purposes of the study were (1) to assess
the mothers’ breast-feeding intentions regarding
their last born infant and to identify variables
associated with their intentions, (2) to assess the
incidence ofbreast-feeding initiation in the hospital
and its associated factors, and (3) to assess the
incidence of breast-feeding practice at home
follow-ing hospital discharge and factors associated with
this practice.
SUBJECTS AND METHODS
The sample consisted of 132 low-income
His-panic mothers participating in the Expanded Food
and Nutrition Education Program in three counties in southern California. From a total of 450 Hispanic mothers enrolled in this program (winter of 1986
and spring of 1987), all mothers of infants 24
months of age or younger were listed as eligible for
the study. Of the 175 mothers eligible, 30 were
excluded because of lack of a home telephone, 10
were not reached by four phone calls, and 3 declined
to participate. The remaining 132 agreed to respond
to the interview and made up the sample of this
study.
The focus of the study was the milk-feeding
intentions before partum and milk-feeding
prac-tices of the mothers regarding their youngest child.
Mothers were asked before their infants were born
whether they had a clear idea as to how they would
feed their infants and what their milk choice was
at that time (breast only, formula only, breast and
formula).
The dependent variables ofthe study were
breast-feeding intentions prepartum, breast-feeding in the
hospital, and breast-feeding at home after hospital
discharge. Feeding modes were defined as follows:
an exclusively breast-fed infant was fed (or
in-tended to be fed) only breast milk with no
supple-mentation of formula. An exclusively formula-fed
infant was fed only formula with no
supplementa-tion of breast milk. A partially breast-fed infant
was fed one or more bottles of formula in addition
to breast milk. In some parts of this article, the
term “total breast-feeding” is used and refers to the
total number of mothers who ever breast-fed,
in-cluding exclusive and partial breast-feeding.
The following sociodemographic information re-lated to the mother and health care provider and the infant was collected. Mother and health care
provider-related data were mother’s age, country of
origin, education, infant’s birth order (first vs
sec-ond or higher), employment during and after
preg-nancy, breast-feeding advise given by health
profes-sionals, type of delivery (cesarean vs vaginal), time
of breast-feeding initiation (hours after birth), type
of milk fed to infant in the hospital, and milk
samples given at the time of hospital discharge.
Infant-related data were birth order, birth weight,
and sex.
Data concerning demographic characteristics
were collected through personal interviews at the
time of enrollment in the Expanded Food and
Nu-trition Education Program. Information regarding
infant feeding was obtained through telephone
in-terviews using Dillman’s techniques’3 and a
stand-ardized survey instrument. Two bilingual Hispanic
students from the University of California at Davis
were trained in telephone interviewing. To increase
reliability, standardization of questioning
tech-nique and recording procedure was performed by
telephone interviewing 20 Hispanic women
partic-ipating in the Expanded Food and Nutrition
Edu-cation Program in a county not included in the
study. Data collection began when recording
van-ation between interviewers decreased to two
incon-sistencies out of 54 recordable items in the
ques-tionnaire.
Data analysis consisted of descriptive, bivaniate,
and multivaniate statistics. Association and
coline-arity among variables were tested using
x2
statisticsand Pearson correlations.’4 Explanation of the
de-pendent variables (breast-feeding intention prepar-tum, breast-feeding in the hospital, and
breast-feeding at home) was assessed using stepwise
logis-tic regression.’5 The role of independent variables
was screened using the forward inclusion method.
Odds ratios for variables entered in the final model
were calculated following Freeman’s method’6 and
using coefficients obtained in the logistic regres-sion.
RESULTS
Sociodemographic
Characteristics
The mean age ofthe mothers was 28.7 ± 5.3 years
and the level of education was 7.8 ± 3.6 years.
Eighty-eight percent of the mothers were from
Mexico and 12.0% were born in the United States. All infants studied were born in US hospitals.
Twenty-eight percent of the mothers worked during
pregnancy and 26.0% returned to work after the
infant was born. Seventy-two percent of the
moth-ens were advised by their doctor or nurse to breast-feed. The average number of children born to these mothers was 3.1 ± 1.5, and the average age of their
youngest child (child studied) was 14.3 ± 3.7
months. Thirty-five of these were firstborn infants and the rest were of higher birth order. Fifty-six
percent were boys. The infants’ mean birth weight
were born at term and 8.0% were born within the
7th and 8th month of pregnancy. Seventy-six
per-cent were born vaginally and 24.0% by cesarean.
The average time of breast-feeding initiation was
13.7 ± 8.0 hours after birth.
Milk-Feeding Intention
As shown in Table 1, choice of feeding mode was
exclusive breast-feeding (67.7%), partial
breast-feeding (10.0%), and exclusive formula (22.3%).
The total number of mothers intending to
breast-feed (exclusive and partial breast-feeding) was 101
(77.7%). The first five sources of influence of their
milk choice reported by mothers were their own
decision independent of others (47.0%), their
mother (10.5%), infant’s father (10.5%), doctor/ nurse (9.0%), and friends/relatives (8.7%).
Cross-tabulations of selected sample
character-istics and milk choice are shown in Table 2.
Inten-tion to breast-feed was significantly associated (P
= .01) with mother’s country of origin. Mothers
TABLE 1. Breast-Feeding Intention Prepartum and
Breast-Feeding Practice in the Hospital and at Home (N = 132)0 ______________________
Prepartum Breast-Feeding Intention
Hospital Home
Breast-feeding mode
Exclusive 67.7 19.7 17.2
Partial 10.0 44.7 46.9
Exclusive formula feeding 22.3 36.2 35.9
0 Results are percentages of women.
from Mexico were more likely to choose
breast-feeding than Hispanic mothers born in the United
States (P = .05). Mothers with 9 or less years of
education were more likely to choose breast-feeding
than mothers who attended school for 10 to 14
years. Logistic regression was used to assess the
independent effect of sociodemographic variables
on the mothers’ intentions to breast-feed. When
the independent variables (mother’s country of
or-igin, age, education, employment, breast-feeding
advice given by health professionals, and infant’s
birth order) were adjusted in the regression model,
the only variable with a significant (P = .05)
influ-ence on milk choice was the mother’s country of
origin (Table 3). The odds of mothers choosing
breast-feeding rather than formula feeding were
4.75 times higher (95% confidence limit 1.80 to
12.50) for women from Mexico than for Hispanic
mothers born in the United States. Mother’s
edu-cation was not significant in the regression analysis.
Milk Fed in the Hospital
As shown in Table 1, the total incidence of
breast-feeding (exclusive and partial
breast-feed-ing) was 63.8%. However, there was a marked
dif-ference between exclusive breast-feeding intention
and practice (Figure ). Although 67.7% of the
mothers intended exclusive breast-feeding
prepar-tum, only about one third (19.7%) exclusively
breast-fed their infants in the hospital. Hospital supplementation of breast milk with formula (par-tial breast-feeding) increased from 10.0% intention
TABLE 2. Sample Characteristics by Mother’s Milk-Feeding Intentions0
Characteristics Milk-Feeding Intention
Breast Formula (n = 101) (n = 29)
Age (y)
18-29 55 (54.5) 18 (62.1)
30-34 46 (45.5) 11 (37.9)
Education ()t
9 70 (69.3) 15 (51.7)
10-14 31 (30.7) 14 (48.3)
Country of origins
Mexico 89 (88.1) 18 (62.1)
United States 12 (11.9) 11 (37.9)
Child’s birth order
First 17 (16.8) 6 (20.7)
Second or higher 84 (83.2) 23 (73.3)
Breast-feeding advise by health professionals
Yes 73 (72.3) 20 (68.9)
No 28 (27.7) 9 (31.1)
Employment during pregnancy
Yes 28 (27.7) 7 (24.0)
No 73 (72.3) 22 (76.0)
0 Results are numbers (%) of women. Breast-feeding includes exclusive and partial
breast-feeding.
tx2 3.07;P .05.
Intention Hospital Home Figure. Exclusive breast-feeding intentions and prac-tice in hospital and at home (N = 132).
TABLE 3. Adjusted Odds Ratio for Breast-Feeding Rather Intention and Practice in the Hospital and at Home0
Than Formula Feeding
Dependent and Independent Variables
Odds Ratio
95% Confidence Limits
Breast-feeding intention (n = 132) Mother’s country of origin
Breast-feeding in the hospital (n = 84)t Time of breast-feeding initiation Milk fed at home (n = 80)t
Time of breast-feeding initiation Infant delivery method
Employment postpartum Country of origin
4.75#{176}
0.96”
1.27” 12.76a 28.26c
854b
1.80-12.50
0.94-0.99
1.04-1.56 1.06-154.08 2.67-229.30 1.17-62.05
0 Adjusted odds ratios were derived from the logistic regression coefficients. Significance:
ap 05;bp .ol;cP= .001.
t
Only the mothers who intended breast-feeding were included to assess factors influencing departure from breast-feeding intention.C U
0.
to 44.7% practice in the hospital. The actual
num-ber of women who exclusively formula fed their
infants in the hospital was almost twice the number
of those who intended this feeding method
prepar-tum. The x2 analysis showed a significant
associa-tion (P = .05) between milk fed in the hospital,
infant delivery method, and time of breast-feeding initiation. Results of the logistic regression analysis showed a variation in the variables impacting breast-feeding initiation. When independent van-ables were adjusted in the regression, the only
van-able having a significant impact on breast-feeding in the hospital was time of breast-feeding initiation
(Table 3). The odds of breast-feeding during the
hospital stay, rather than exclusive formula feeding, were higher for women who initiated breast-feeding
within 10 hours after birth than for those who
initiated breast-feeding 11 or more hours after birth (odds ratio 0.96, 95.0% confidence limits 0.94 to 0.99). Child’s sex, birth order, and birth weight had no impact on type of milk fed in the hospital.
Milk Fed at Home
There was a large difference between
milk-feed-ing intentions prepartum and milk fed at home
after hospital discharge. Although 67.7% of mothers
intended exclusive breast-feeding, only one fourth
of them exclusively breast-fed their infants at
home. Correspondingly, the number of mothers who
supplemented breast-feeding with formula-feeding
increased by 4.5 times. The number of mothers who
exclusively formula fed increased by 1.5 times.
Re-sults of
x2
analysis showed a significant association(P = .05) between milk fed at home and delivery
method, formula samples given at the time of
hos-pital discharge, mother’s country of origin, and
mother’s employment postpartum. When logistic
regression was used to assess the impact of the
independent variables on milk fed at home, the
following variables demonstrated a significant
in-fluence on this practice: time of breast-feeding
mi-tiation, delivery method, mother’s employment
postpartum, and mother’s country of origin (Table
3). Consequently, the odds for breast-feeding at
home were 1.27 greater for mothers who initiated
breast-feeding in the hospital within the first 10
hours rather than 11 hours or later, 12.76 times
greater for mothers who had a vaginal delivery as opposed to cesanean section, 28.26 times greater for mothers not returning to work after pregnancy, and 8.54 times greater for mothers from Mexico than for mothers born and raised in the United States. Although there was an association in the
x2
analysis between formula samples given at the time ofhos-pital discharge and formula feeding at home, it did
not reach statistical significance in the regression
analysis. Mother’s education and other
demo-graphic variables related to the infant and mother did not have a significant role in breast-feeding practice at home.
DISCUSSION
Our results for total breast-feeding intentions,
77.7% (exclusive and partial breast-feeding), are
much higher than the 22.6% reported by Rassin et
a14 from Texas and are similar to those reported by Scrimshaw et al7 for a similar group of Hispanic
mothers in Los Angeles. These authors found that
82.0% of mothers in one of the two hospitals they
studied wanted to breast-feed. However, a
break-down of figures for exclusive and partial
breast-feeding are not provided. Because of the negative
impact of formula supplementation on breast-feed-ing initiation and dunation,5’6”72#{176} the distinction
between these two breast-feeding modes is
impor-tant. Our figures for exclusive breast-feeding
inten-tion prepartum and for partial breast-feeding were
67.7% and 10.0%, respectively. Breast-feeding
in-tentions were significantly associated with mother’s
country of origin. Mothers from Mexico were
sig-nificantly (P = .01) more likely to intend
breast-feeding than mothers born in the United States,
suggesting a tendency toward the preservation of
traditional infant-feeding practices prevailing in
many areas of Mexico.21’22
In our study, there was a large discrepancy
be-tween exclusive breast-feeding intentions and
prac-tice. Although 67.7% of the mothers intended to
exclusively breast-feed, only about one fourth of
them (19.7%) did so in the hospital and 17.2% at
their home. Our figure for total breast-feeding
prac-tice (exclusive and partial breast-feeding) is 64.1% which is similar to the 75.0% breast-feeding goal
stated in the Surgeon General’s report23 for the
United States in 1990. Our data are also similar to
the 60.0% reported by Samuel et a!6 among
His-panic mothers in Oakland and by Young et al8 for
mothers of Mexican descent in North Carolina.
However, our breast-feeding rate is much higher than that reported by Smith et al3 for the subsample
of mothers studied between 1976 and 1979 in 51
counties along the United States/Mexican border.
This inconsistency may be due to differences in the
samples studied. The mothers in our sample were all below the poverty level (requirement for
enroll-ment in the Expanded Food and Nutrition
Educa-tion Program program) and levels of education were
low. The mothers in the Smith et a!3 sample had
higher levels of education (33.4% had 13 or more
years of schooling as opposed to 10.0% in our
sam-ple). Smith et a13 did not provide data regarding
income and did not distinguish between migrants
and mothers born in the United States.
It is not clear whether the extensive
formula-feeding practice in the hospital (80.9% total: 36.2%
exclusive formula feeding and 44.7% formula
sup-plementing breast) was the primary cause for the
sharp decline from exclusive breast-feeding
inten-tion to practice. Our data document type of milk
fed in the hospital, but it is not known whether the
formula was provided at the mother’s
request/con-sent or whether mothers misinterpreted formula
bottles in their infant’s bassinet as a prescription from their health care providers. Including formula
bottles in the bassinets of infants is a common
practice in many hospitals in California. Infants
whose mothers intend to exclusively breast-feed are
not always identified, and their bassinets are not
always labeled to prevent inclusion of formula bot-tles.
The time of breast-feeding initiation had a
sig-nificant impact on the rate of breast-feeding.
Moth-ers who attempted breast-feeding within the first
10 hours after birth were more likely to continue
breast-feeding during the hospital stay than
moth-ers who attempted breast-feeding 11 hours or more
after birth. Our results are consistent with those of
Feinstein et al’7 among low income mothers in
Chicago, who found that the incidence of
breast-feeding was reduced when the onset of breast-feed-ing was 16 or more hours postpartum. Scrimshaw
et al,7 Salaniya et al,24 and Taylor et al25 also
re-ported early breast-feeding onset as a significant
factor in breast-feeding success. Well-known pedia-tnicians specializing in breast-feeding6’27
recom-mend initiating breast-feeding as soon as possible
after delivery. Clinicians indicate that the period following birth when infants are most alert is the
optimum time to initiate proper “latch on” and
suckling behavior.25’26 Early and frequent suckling
contributes to the prevention of milk stasis in the
duct system and sinuses, consequently contributing
to the prevention of engorgement. Unrelieved
en-gorgement can inhibit milk production by causing
pressure and atrophy of milk secretory cells.28’
The feeding mode practiced in the hospital was
almost replicated at home (Table 1). It appears that
the hospital experience determined the feeding
mode for the infant afterward. Mothers who
for-mula fed in the hospital continued formula feeding
at home. Only two of the mothers who did not
initiate breast-feeding in the hospital attempted
breast-feeding at home.
The type of delivery method influenced the rate
of breast-feeding. Mothers whose infants were
de-livened by cesanean section were 12.76 times less
likely to breast-feed than mothers whose babies were delivered vaginally. These results are consist-ent with other studies.6’7’3#{176}Samuels et a!6 reported a breast-feeding incidence of 52.0% and 69.0% for cesarean section and vaginal births, respectively.
The rate of cesanean births, however, in their
mul-tiethnic sample was 18.0% compared to the 24.1%
higher than the 18.8% average rate for births in
California.3’ Although the discomfort caused by
cesarean section deliveries can be a deterrent to
breast-feeding, techniques developed by Frantz and
Kalmen32 have been effective in making
breast-feeding possible. In our study, only 2 of the 32
mothers attempted to breast-feed after cesarean
section.
The negative impact of mother’s employment
postpartum on breast-feeding in our study is
con-sistent with data obtained by Martinez and Stahle33
among mothers participating in the Supplemental
Food Program for Women, Infants, and Children.
In our sample, 28.0% of the mothers were employed
4 to 10 h/d during pregnancy and 26.0% worked
this much postpartum. The odds of breast-feeding
for women in our sample who did not return to
work after delivery were 28.6 times higher than for
women who returned to work. The study by
Mar-tinez and Stahle33 is one of the few in which the
effect of maternal employment on breast-feeding
among low income minorities was reported. Their
data indicated that mothers who were employed
full-time had lower rates of breast-feeding
com-pared with mothers who were employed part-time
postpartum. The duration of breast-feeding among
women who did not have plans to work was twice
as high as for employed women. The limited
avail-able studies suggest that, whereas mother’s
employ-ment can be significantly detrimental to
breast-feeding among the poor, it does not appear to have
the same impact on women of higher education and
income levels.33’34
Contrary to many reports in the literature,
moth-er’s education did not have a positive impact on
breast-feeding. Its influence tended to be in the
opposite direction. The least educated women were
most likely to breast-feed. These findings are, in
part, consistent with the findings of Rassin et al4
among Hispanic women in Texas. Their results
imply a U-shaped distribution in that the highest
breast-feeding rate was among the least and the
most educated mothers. Women with education
in-between the two extremes were the least likely to breast-feed.
To summarize, Hispanic mothers in our study
had breast-feeding intentions slightly greater than
the 75.0% goal established by the Surgeon General’s
Workshop on Lactation and Breast-Feeding for
American women for the year 1990. Although the
total number of mothers in our sample who even
breast-fed their infants did not decrease drastically,
the incidence of exclusive breast-feeding practice
was about 4 times less than intended. Both
exclu-sive formula feeding and partial breast-feeding
in-creased significantly from prepartum intention to
practice. Factors associated with breast-feeding
de-dine were specific hospital protocols and practices, as well as mother’s employment.
Intervention and policy implications of this study
include the development/improvement of hospital
practices supportive of breast-feeding, such as (1) early initiation of breast-feeding after delivery; (2)
special assistance and encouragement to mothers
whose babies are delivered by cesarean section and
wish to breast-feed; (3) instruction to employed
mothers regarding extraction, storage, and use of
stoned breast-milk; (4) training in the clinical as-pects of breast-feeding for hospital staff in obstet-nics services; and (5) avoidance of in-hospital for-mula supplementation to breast-fed infants (unless medically justified) and avoidance of formula dis-tnibution at the time of hospital discharge. The negative health and economic implications of for-mula use can be significant, particularly among
families with limited income and/on inadequate
home sanitation. The Surgeon General’s Workshop on Breast-Feeding and Human Lactation23 and the WHO International Code of Marketing of Breast-milk Substitutes35 articles 6.1 to 6.8 cleanly outline recommendations for health services to adopt.
Finally, the retrospective nature of this study
should be noted when interpreting our results.
Per-spective follow-up studies are needed to control for possible recall bias. More research concerning spe-cific ethnic minorities is needed to guide policy
making and programs that target populations at
risk. To prevent undennepresentation of minority
groups with limited education, written
question-names (the data collection method used in most
breast-feeding studies) should be avoided.
ACKNOWLEDGMENTS
We thank Mary Marshall, MA, Sandra Spencer, MA, and Eunice Williamson, MS, Expanded Food and Nutni-tion Education Program (EFNEP) coordinators in San Bernardino, Los Angeles, and Riverside counties for their cooperation in this study. We also thank Edward Dolber-Smith, BS, for his assistance in data analysis, Denise DiPietro, BA, for her assistance in editing the manu-script, and Dr Amy Block Joy, Director of the California EFNEP for the support provided for data collection and analysis.
REFERENCES
1. Martinez GA, Krieger FW. 1984 milk feeding patterns in the United States. Pediatrics. 1985;76:1004-1008
2. Faden RB, Gielen A. Contemporary breast-feeding patterns: focus on disadvantaged women. Clin Nutr. 1986;5:200-209 3. Smith JC, Mhango CG, Warren CW, et al. Trend in the
incidence of breast-feeding for Hispanics of Mexican origin and Anglos on the USA-Mexican border. Am J Public
Health. 1982;72:59-61
4. Rassin DK, Richardson CJ, Baranowski T, et al. Incidence of breast-feeding in a low socioeconomic group of mothers
132-137
5. Kokinos M, Dewey KG. Infant feeding practices of migrant Mexican-American families in Northern California. Ecol
Food Nutr. 1986;18:209-220
6. Samuels SE, Margen S, Schoen EJ. Incidence and duration of breast-feeding in a health maintenance organization pop-ulation. Am J Clin Nutr. 1985;44:504-510
7. Scrimshaw SCM, Engle PL, Arnold L, et al. Factors affect-ing breast-feedaffect-ing among women of Mexican origin or de-scent in Los Angeles. Am J Public Health. 1987;74:467-470 8. Young SA, Kaufman M. Promoting breast-feeding at a
migrant health center. Am J Public Health. 1988;78:523-525 9. Mohrer J. Breast and bottle feeding in an inner city
com-munity: an assessment of perception and practices. Med
Anthrol. 1979;39:125-145
10. Bryant CA. The impact ofkin, friend and neighbor networks on infant feeding practices. Soc Sci Med. 1989;16:1757-1765 11. Serger MT, Gibbs CE, Young EA. Attitudes about
breast-feeding in a group of Mexican-American primigravidas.
Texas Med. 1979;75:78-80
12. Wittember CK. Summary of market research for healthy mothers, healthy babies campaign. Public Health Rep.
1983;98:356-359
13. Dillman DA. Mail and Telephone Survey. New York, NY: John Wiley & Sons; 1978:85 14 SPSS, 2nd ed. New York, NY: McGraw-Hill Book Co; 1986:336
15. Engelman L. Stepwise logistic regression. In Dixon WJ, Brown MB, Engleman L, et al, eds. BMDP Statistical
Soft-ware. Berkeley, CA: University of California Press; 1985:330
16. Freedman D Jr. Applied Categorical Data Analysis. New York, NY: Marcel Dekker Inc; 1987:242
17. Feinstein JM, Berkelhamer JE, Gruszka ME, Wong CA, Carey AE. Factors related to early termination of breast-feeding in an urban population. Pediatrics. 1986;78:210-215 18. Kurinij N, Shiono PH, Rhoads GG. Breast-feeding incidence
and duration among black and white women. Pediatrics.
1988;81:365-371
19. de Chateau P. A study of factors promoting and inhibiting lactation. Dev Med Child Neurol. 1977;19:575-584
20. Winikoff B, Laukaran VH, Myers D, Stone R. Dynamics of infant feeding: mothers, professionals, and the institutional context in a large urban hospital. Pediatrics. 1986;77:357-365
21. Segunda Encuesta Nacional de Alimentacion-1979: La Al-imentacion en el Medio Rural, Mexico. Mexico City, Mexico: Instituto Nacional de Nutricion; 1980
22. Romero-Gwynn E. Family Well-being, Fertility and Child Nutrition: A Comparative Study Between Migrant and Na-tive Families in Guadalajara, Mexico. Ithaca, NY: Cornell University; 1977. PhD dissertation23 Report of the Surgeon General’s Workshop on Breast-Feeding and Human
Lacta-tion. Washington, DC: Government Printing Office; 1984. US Dept of Health and Human Services publication HRS-D-MC 84-2
24. Salariya EM, Easton PM, Carter JI. Duration of breast-feeding after early initiation and frequent feeding. Lancet.
1978;2:1114-1143
25. Taylor PM, Maloni JA, Brown DR. Early suckling and prolonged breast-feeding. Am J Di.s Child. 1986;140:151-154 26. Neifert RM, Seacat J. A guide to successful breast-feeding.
Contemp Pediatr. 1986;3:1-14
27. Lawrence R. Breast-Feeding: A Guidefor the
MedicalProfes-sian. St Louis, MO: CV Mosby Co; 1985:186
28. Neifert MR. Routine management of breast-feeding. In Ne-ville MC, Neifert MR, eds. Lactation Physiology, Nutrition
and Breast-feeding. New York, NY: Plenum Press; 1983:284
29. Fleet IR, Peaker M. Mammary function and its control at the cessation of lactation in the goat. J Physiol. 1978;
279:491-507
30. Palmer RS, Arvey A, Taylor R. The influence of obstetric procedures and social and cultural factors on breast-feeding rates at discharge from hospital. J Epidemiol Community
Health. 1979;33:248-252
31. Oreglia A. Health Data Summaries for California Countries:
1984 Report. Sacramento, CA: Center for Health Statistics,
Department of Health Services, January 1984
32. Frantz FB, Kalmen BA. Breast-feeding works for cesareans too. RN. 1979;39-47
33. Martinez G, Stahle DA. The recent trend in milk feeding
among WIC infants. Am J Public Health. 1982;72:68-71 34. Auerbach K, Guss E. Maternal employment and
breast-feeding: a study of567 women’s experiences. Am J Dis Child.
1984;138:958-960
35. International Code of Marketing of Breastmilk Substitutes.