You are on page 1of 8

679384

research-article2016
JHLXXX10.1177/0890334416679384Journal of Human LactationJefferson

Original Research
Journal of Human Lactation

Breastfeeding Exposure, Attitudes, and


2017, Vol. 33(1) 149­–156
© The Author(s) 2016
Reprints and permissions:
Intentions of African American and sagepub.com/journalsPermissions.nav
DOI: 10.1177/0890334416679384

Caucasian College Students journals.sagepub.com/home/jhl

Urmeka T. Jefferson, PhD, RN1

Abstract
Background: African American mothers lag behind in breastfeeding initiation. Research is needed to gain an understanding
of potential reasons for breastfeeding disparities.
Research aim: This study explored breastfeeding exposure, attitudes, and intentions of African American and Caucasian
college students by race and gender.
Methods: Women and men (696) attending college, who were younger than 45 years and without children, were included
in this study. Survey data were collected using a demographic questionnaire and the Iowa Infant Feeding Attitude Scale.
Results: Overall, students demonstrated favorable attitudes regarding breastfeeding but viewed formula feeding as
more practical. Students who were Caucasian and female and experienced breastfeeding exposure demonstrated higher
breastfeeding attitudes and intent. Breastfeeding exposure and attitudes contributed 32% of the variance in breastfeeding
intentions. The odds of experiencing breastfeeding exposure and positive breastfeeding attitudes were approximately 3 times
higher for Caucasian students than for African American students.
Conclusion: External factors demonstrated a stronger association with breastfeeding intentions. The link with race and
gender appears to operate through their effect on attitudes and exposure. More research is needed to identify strategies to
improve breastfeeding exposure and attitudes among African Americans.

Keywords
African Americans, breastfeeding, breastfeeding rates, health status disparities, human milk

Breastfeeding is the most protective measure to improve breastfeeding choices (Street & Lewallen, 2013). Breast-
infant health and mortality. Research has demonstrated a feeding exposure, attitudes, knowledge of breastfeeding
significant reduction in the risk of infections (respiratory, benefits, and the complex lifestyles of working women were
ear, and gastrointestinal), asthma, sudden infant death syn- identified as factors external to demographic variables that
drome, and necrotizing enterocolitis for breastfed infants (Ip determine breastfeeding outcomes for African American
et al., 2007). Therefore, the American Academy of Pediatrics women (Jefferson, 2014a).
recommends exclusive breastfeeding for the first 6 months Breastfeeding initiation and duration among all races
after birth (Eidelman et al., 2012). Breastfeeding rates in have been associated with specific demographic characteris-
the United States have seen an increase from 71% to 81% tics. A young maternal age, low income, less education, and
between 2002 and 2013, nearing the Healthy People 2020 unmarried status have been commonly linked with low
national breastfeeding goal of 81.9% for infants receiving breastfeeding rates (CDC, 2010, 2016). However, college-
any human milk. Currently, only Caucasian (84%), Asian educated African American women do not initiate breast-
(84%), and Hispanic (83%) mothers have reached the feeding at the same rate as college-educated Caucasian
Healthy People 2020 goal, and African American mothers women (CDC, 2010). This racial disparity in breastfeeding is
(66%) have the lowest rate for any breastfeeding (Centers
for Disease Control and Prevention [CDC], 2016). Previ-
1
ous research has shown that greater comfort with formula Sinclair School of Nursing, University of Missouri, Columbia, MO, USA
feeding accounts for the racial difference in breastfeeding Date submitted: July 25, 2016; Date accepted: October 22, 2016.
intentions (Nommsen-Rivers, Chantry, Cohen, & Dewey,
Corresponding Author:
2010). In addition, African American women have reported Urmeka T. Jefferson, PhD, RN, Sinclair School of Nursing, University of
that culture did not influence their breastfeeding decision, Missouri, S416 School of Nursing, Columbia, MO 65211, USA.
indicating that external factors may have a greater effect on Email: jeffersonu@missouri.edu
150 Journal of Human Lactation 33(1)

not explicated by demographic characteristics (e.g., educa-


tion) alone, signifying that external factors within each race Key Messages
may have a stronger association with breastfeeding. Limited •• Despite breastfeeding promotion, African
research has examined breastfeeding beliefs among popula- American women have lower breastfeeding rates
tions with a higher level of education beyond high school. than other women.
Since 2000, studies in the United States that explored infant- •• The racial disparity is not explained by demo-
feeding attitudes, knowledge, exposure, experiences, and graphic variables alone, indicating that external
intentions of college students have appeared in peer-reviewed factors may have a greater effect on breastfeeding.
journals. Evidence demonstrates that breastfeeding expo- •• Breastfeeding exposure and attitudes had a stron-
sure, knowledge, and attitudes predicted breastfeeding inten- ger association with intentions. More Caucasians
tions (Dodgson, Bloomfield, & Choi, 2014; Froehlich et al., experienced exposure and positive attitudes than
2013; Jefferson, 2014b, 2015; Kavanagh, Lou, Nicklas, African Americans.
Habibi, & Murphy, 2012; Marrone, Vogeltanz-Holm, & •• Innovative interventions to increase breastfeeding
Holm, 2008; Spear, 2006; Vari et al., 2013). Although the exposure and attitudes among African Americans
evidence among college-educated students in the United are needed.
States provides a view of breastfeeding beliefs, the samples
were primarily Caucasian students all from one course or
major (Dodgson et al., 2014; Kavanagh et al., 2012; Marrone among this population is low and Caucasian students were
et al., 2008; Spear, 2006) and one exclusive sample of used as a representation for the racial group with the highest
African American students (Jefferson, 2014b, 2015), with no breastfeeding rates. To ensure recruitment of group samples
comparisons made among racial groups. large enough to explore race and gender differences in breast-
Research continues to move beyond identifying demo- feeding exposure, attitudes, and intentions among African
graphic characteristics as a single indicator that predicts Americans and Caucasians, a power analysis for indepen-
breastfeeding initiation and duration. Limited focus has been dent-samples t-tests using a .05 level of significance, an effect
on breastfeeding beliefs and intentions prior to pregnancy. size of 0.30, and 0.80 power indicated that approximately 174
Evidence has demonstrated that beliefs regarding breastfeed- students per race and gender group were needed. The total
ing are formed before becoming pregnant (Jefferson, 2014b; sample provided sufficient cases (20 to 1) per variable for
Seidel, Schetzina, Freeman, Coulter, & Colgrove, 2013). As regression analysis to assess predictors of breastfeeding
a result, research with emphasis on breastfeeding determi- intentions (Polit & Beck, 2004).
nants prior to pregnancy is necessary to identify potential
factors associated with breastfeeding intentions for racial
groups with the lowest breastfeeding rates. It is also essential Data Collection and Measures
to identify cultural and social factors that influence breast-
Recruitment occurred in various locations on campus from
feeding attitudes and decisions affecting low breastfeeding
February to October 2013 by distributing flyers in the stu-
rates of minority women (Jones, Power, Queenan, &
dent center and during campus events at data collection
Schulkin, 2015). The specific aims of this study among a
times. The purpose, risks, and benefits were explained in a
sample of college students without children were to (1)
recruitment statement letter. Participation was voluntary and
explore race and gender differences in breastfeeding expo-
completion of questionnaires inferred consent. Students
sure and attitudes, as well as breastfeeding intentions, and (2)
from all majors were invited to participate in the study. Name
assess the contribution of race, gender, breastfeeding expo-
and student number was collected from each participant and
sure, and attitudes to breastfeeding intentions.
verified with the student identification card at enrollment.
Participant information was kept on an enrollment list on a
Methods password-protected laptop computer to prevent duplicate
participation. The list was checked for participation before
Design and Sample questionnaires were provided. Participants completed the
Institutional review board approval was received from the Iowa Infant Feeding Attitude Scale (IIFAS) and demographic
University of Missouri for this descriptive cross-sectional questionnaire in paper format at the time of enrollment. The
study. A total sample of 696 college students was included IIFAS collected data about attitudes toward breastfeeding
based on the following criteria: women and men attending the and formula feeding. The demographic questionnaire col-
University of Missouri–Columbia; self-identified as African lected data regarding demographic information, breastfeed-
American or Caucasian; younger than 45 years; able to speak, ing exposure, and intentions. Each student received a $5 gift
read, and write English; do not have any children; and no his- card for participation. The risk of loss of anonymity was
tory of pregnancies. Although Asians and Hispanics have minimized by not recording names on the questionnaires.
high national rates of breastfeeding, on-campus enrollment Questionnaires were not linked to the enrollment list.
Jefferson 151

Iowa Infant Feeding Attitude Scale.  The IIFAS is a 5-point Lik- demographic data, breastfeeding exposure, attitude scores,
ert-type scale developed to measure infant-feeding attitudes. and breastfeeding intention. Differences in attitude scores
Items were developed regarding parental role, costs, nutri- were analyzed using independent-samples t-tests. Chi-square
tion, convenience, and infant bonding, with responses rang- test of independence was run to explore associations of
ing from 1 (strongly disagree) to 5 (strongly agree). Final breastfeeding exposure, intentions, race, and gender. Logistic
items were selected based on rank ordered item correlations regression was used to assess the contribution of race, gen-
resulting in a 17-item scale (de la Mora, Russell, Dungy, der, breastfeeding exposure, and attitudes to breastfeeding
Losch, & Dusdieker, 1999). Eight items reference breast- intentions.
feeding, whereas nine items reference formula feeding. After
formula-feeding items were reverse coded, the score was
totaled, providing a range of 17 to 85. Positive breastfeeding
Results
attitudes are represented by a score approaching 85, and pos- Data were collected from 696 students with a mean (standard
itive formula-feeding attitudes are represented by a score deviation [SD]) age of 20 (2.49) years (range = 18-44) and
near 17 (de la Mora et al., 1999). The IIFAS has been used to mean (SD) education of 14 (1.24) years. Table 1 reports
predict feeding decisions, differentiating breastfeeding from demographic characteristics, breastfeeding exposure, and
formula-feeding mothers. Phrasing of items allows use with intentions of the sample by gender.
various individuals, rather than being constrained to mothers
or pregnant women (Chambers, McInnes, Hoddinott, &
Alder, 2007). Mean attitude scores have been analyzed for
Breastfeeding Exposure
group comparisons of infant-feeding intentions and behav- There was a significant association between being breastfed
ior, with breastfeeding women found to have significantly as an infant and knowing someone who breastfed her infant
higher attitude scores ranging from 63 to 66 compared with (χ2 = 36.26, p < .001). Of the students reporting that they
scores of 49 to 56 for women who formula feed their infant were breastfed as an infant, 97% also knew someone who
(de la Mora et al., 1999; Shaker, Scott, & Reid, 2004; Sitt- breastfed her infant. A chi-square test for independence indi-
lington, Stewart-Knox, Wright, Bradbury, & Scott, 2007). cated a significant association between being breastfed as an
The IIFAS demonstrated adequate internal consistency of a infant and Caucasian race (χ2 = 24.93, p < .001) but not gen-
Cronbach’s alpha of .68 to .86, demonstrating reliability in der (p = .162). A significant association was identified
assessing infant-feeding attitudes as well as validity reflected between knowing someone who has breastfed and Caucasian
by the ability to predict breastfeeding initiation and duration race (χ2 = 10.27, p = .001) as well as female gender (χ2 =
(de la Mora et al., 1999). Additional studies using the IIFAS 7.01, p = .008). Table 2 presents race differences within gen-
have demonstrated acceptable internal consistency with a der groups.
Cronbach’s alpha of .79 (Shaker et al., 2004; Sittlington
et al., 2007).
Breastfeeding Attitudes
Demographic questionnaire. The demographic questionnaire Total sample scores ranged from 29 to 82 with a mean (SD)
was constructed for this study to gather data pertaining to attitude score of 59 (7.47), and the Cronbach’s alpha coeffi-
demographic variables characteristic of infant-feeding deci- cient was .72. Cronbach’s alpha by race was .70 for African
sions as well as breastfeeding exposure and intentions. Americans and .75 for Caucasians. Table 3 displays total
Demographic questions included age, gender, race, marital sample item responses. More male students (45%) agreed
status, student status, education level, income, work status, that formula feeding is a better option if the mother works
and state of residence. Breastfeeding exposure was measured outside the home compared with females (28%). Female stu-
by report of being breastfed as an infant and knowing some- dents (54%) also disagreed with the statement that formula is
one other than your mother who breastfed her infant. Breast- as healthy as human milk, compared with 42% of males.
feeding intention was measured by one question asking There was a significant difference in attitude scores by race,
students to identify the probability of breastfeeding your gender, being breastfed as an infant (the “don’t know” group
future child or providing support to your wife/girlfriend in was excluded), and knowing someone who has breastfed
breastfeeding your future child that used a 6-point Likert- (see Table 4).
type scale, with responses ranging from no chance to 81%-
100% probability.
Breastfeeding Intentions
Given the small numbers in the lower categories (see Table 1),
Data Analysis breastfeeding intentions were collapsed into two categories
Data were analyzed using SPSS Statistics 21. Internal con- of highly likely (> 80% probability) and less likely (≤ 80%
sistency reliability using Cronbach’s alpha was estimated for probability) for subsequent analyses. A significant associa-
the IIFAS. Descriptive statistics were used to delineate the tion was noted between breastfeeding intentions and race,
152 Journal of Human Lactation 33(1)

Table 1.  Demographic Characteristics of Participants Logistic regression was used to assess the contribution of
race, gender, exposure, and attitudes to breastfeeding inten-
Characteristic Total Female Male
tions. For the numeric variable attitudes, the assumption of a
Race linear relationship between attitudes and the log-odds of the
  African American 348 (50) 174 (50) 174 (50) response variable was found to be reasonable. The first
 Caucasian 348 (50) 174 (50) 174 (50) model included only race and gender. This model was statis-
Marital statusa tically significant (χ2 = 29.143, p < .001; Nagelkerke R2 =
 Single 687 (99) 344 (99) 343 (99) .055) and correctly classified 62% (based on a cutoff point
 Married 4 (0.6) 1 (0.3) 3 (0.8) of 0.5) of the cases. Race (p < .001) and gender (p < .05)
 Separated 2 (0.3) 1 (0.3) 1 (0.3) contributed significantly to the model. The full model (see
 Divorced 1 (0.1) 1 (0.3) 0 (0) Table 6) included all variables and demonstrated statistical
Student status
significance (χ2 = 223.502, p < .001; Nagelkerke R2 = .371).
 Full-time 682 (98) 339 (97) 343 (99)
The full model correctly classified 75% (based on a cutoff
 Part-time 14 (2) 9 (3) 5 (1)
point of 0.5) of the cases. Attitudes, being breastfed as an
Education level
infant, and knowing someone who breastfed were the only
 Undergraduate 661 (95) 328 (94) 333 (96)
 Graduate 35 (5) 20 (6) 15 (4)
variables that significantly contributed 32% of the variance
Work statusa in high probability to breastfeed. As a result, additional anal-
 Full-time 30 (4) 13 (4) 17 (5) yses were conducted to examine the effect of race and gender
 Part-time 419 (60) 230 (66) 189 (55) on breastfeeding exposure and attitudes. Separate logistic
  Do not work 244 (35) 105 (30) 139 (40) regression models were conducted to assess the ability of
Yearly incomea (in US$) race and gender to predict exposure by knowing someone
 0-14,999 648 (93) 329 (95) 319 (92) who breastfed and being breastfed as an infant, whereas a
 15,000-29,999 36 (5) 15 (4) 21 (6) multiple regression model was conducted for attitudes (see
 30,000-44,999 9 (1) 3 (1) 6 (2) Table 7). Race and gender were the only predictor variables
 45,000-59,999 0 (0) 0 (0) 0 (0) entered in each model. Students responding “don’t know” (n
 60,000+ 2 (0.3) 0 (0) 2 (0.6) = 142) regarding being breastfeed as an infant were excluded
Parental financial supporta from that model. The odds for Caucasians knowing someone
 Yes 554 (80) 272 (78) 282 (81) who breastfed and being breastfed as an infant are approxi-
Parent yearly income (in US$) mately 2.5 times higher than for African Americans. Whereas
 0-14,999 23 (3) 14 (4) 9 (3) the odds of being breastfed were the same for males and
 15,000-29,999 72 (10) 42 (12) 30 (9) females, with no significant contribution of gender to the
 30,000-44,999 76 (11) 43 (12) 33 (9) model, the odds of knowing someone who breastfed were
 45,000-59,999 74 (11) 40 (12) 34 (10) about 2 times higher for females than males. Attitude scores
 60,000+ 302 (43) 142 (41) 160 (46) were also predictable from race and gender. Caucasians and
Breastfed as an infanta 388 (56) 205 (59) 183 (53)
females scored about 3 points higher on the IIFAS. Based on
Know someone who has 640 (92) 330 (95) 310 (89)
the odds ratios and regression coefficients, race was the
breastfed
Breastfeeding intention
strongest predictor for both exposure and attitudes.
  No chance 14 (2) 7 (2) 7 (2)
  1%-20% probability 11 (2) 5 (1) 6 (2) Discussion
  21%-40% probability 31 (4) 15 (4) 16 (5)
  41%-60% probability 89 (13) 32 (9) 57 (16) This study used a large sample of college students with
  61%-80% probability 142 (20) 66 (19) 76 (22) women and men of both African American and Caucasian
  81%-100% probability 409 (59) 223 (64) 186 (53) race, allowing statistical analysis that explored differences in
breastfeeding exposure, attitudes, and intentions. This evi-
Note. N = 696. Data are given as n (%). Totals may not equal 100% due to dence contributes to the existing body of knowledge by gain-
rounding and missing responses for some questions.
a
Missing values: marital status = 2; work status = 3; yearly income = 1;
ing a better understanding of external factors that may
parental financial support = 3; breastfed as an infant = 2. contribute to the disparity in breastfeeding. Although race
and gender were significant contributors to breastfeeding
intentions when considered alone in this study, they were not
gender, being breastfed as an infant, and knowing someone significant in the full model. When all variables were consid-
who has breastfed (see Table 5). A significant difference in ered, breastfeeding exposure and attitudes contributed the
attitude scores was observed among breastfeeding intentions highest variance in breastfeeding intentions, indicating a
(t = −12.85, p < .001). Students highly likely to breastfeed stronger association with intentions than race and gender.
had higher attitude scores (M = 62, SD = 7.01), whereas stu- The link of race and gender with breastfeeding intentions
dents less likely had a mean (SD) attitude score of 55 (6.35). appears to work through their effect on exposure and
Jefferson 153

Table 2.  Race Differences Within Gender Groups

Female (n = 348) Male (n = 348)

African African
American Caucasian American Caucasian

Characteristic n (%) n (%) χ2 p n (%) n (%) χ2 p


Breastfeeding exposure
  Breastfed as an infant 86 (50) 119 (68) 15.76 < .001 80 (46) 103 (59) 8.99 .01
  Know someone who breastfed 161 (92) 169 (97) 2.87 .09 147 (84) 163 (94) 6.65 .01
Breastfeeding intentions 6.04 .01 −14.15 < .001
  Highly likely 100 (58) 123 (71) 75 (43) 111 (64)  
  Less likely 74 (42) 51 (29) 99 (57) 63 (36)  

Table 3.  Total Sample Iowa Infant Feeding Attitude Scale Responses

Item Agree Neutral Disagree


a
The nutritional benefits of breast milk last only as long as the infant is breastfed. 170 (24) 139 (20) 387 (56)
Formula feeding is more convenient than breastfeeding.a 334 (48) 160 (23) 202 (29)
Breastfeeding increases mother–infant bonding. 625 (90) 54 (8) 17 (2)
Breast milk is lacking in iron.a 35 (5) 372 (53) 289 (42)
Formula-fed infants are more likely to be overfed than are breastfed infants. 295 (42) 257 (37) 144 (21)
Formula feeding is the better choice if a mother plans to work outside the home.a 255 (37) 171 (24) 270 (39)
Mothers who formula feed miss one of the great joys of motherhood. 313 (45) 207 (30) 176 (25)
Women should not breastfeed in public places such as restaurants.a 269 (39) 122 (17) 305 (44)
Infants fed breast milk are healthier than infants who are fed formula. 453 (65) 174 (25) 69 (10)
Breastfed infants are more likely to be overfed than formula-fed infants.a 65 (9) 251 (36) 380 (55)
Fathers feel left out if a mother breastfeeds.a 69 (10) 107 (15) 520 (75)
Breast milk is the ideal food for infants. 547 (79) 110 (16) 39 (5)
Breast milk is more easily digested than formula. 413 (59) 242 (35) 41 (6)
Formula is as healthy for an infant as breast milk.a 149 (21) 212 (31) 335 (48)
Breastfeeding is more convenient than formula feeding. 232 (33) 181 (26) 283 (41)
Breast milk is less expensive than formula. 633 (91) 40 (6) 23 (3)
A mother who occasionally drinks alcohol should not breastfeed her infant.a 423 (61) 148 (21) 125 (18)

Note. N = 696. Data are given as n (%). Responses were collapsed to include three categories: agree (strongly agree and agree), disagree (strongly disagree
and agree), and neutral.
a
This indicates an item in favor of formula feeding.

attitudes. Jefferson (2014b) also found attitudes to be the attitude scores than males. Although this finding was similar
strongest predictor, whereas Marrone and colleagues (2008) to a study with a total sample of African American students
found knowledge scores to be the only significant contribu- (Jefferson, 2014b), it contrasted with studies that found no
tor in predicting intentions to breastfeed. significant difference in breastfeeding attitudes by gender
Although a statistically significant association was (Dodgson et al., 2014; Marrone et al., 2008). Including men
observed between infant-feeding attitudes, race, gender, and in this sample permitted data collection directly from a
breastfeeding exposure, less than or equal to a 5-point differ- male’s point of view regarding infant-feeding beliefs instead
ence was noted between the mean attitude scores among of relying only on women’s perspectives of their partner’s
groups. This significant change indicates that even a small view.
improvement in breastfeeding attitudes and exposure for Similar to other findings (Jefferson, 2014b; Kavanagh
African Americans could help eliminate the disparity in et al., 2012), overall, students agreed that breastfeeding
breastfeeding rates. Data showed that the odds of experienc- improves mother–infant bonding, costs less, and is healthier
ing breastfeeding exposure and higher attitude scores were and that nutrition benefits continue after weaning. Formula
greater for Caucasians than for African Americans, demon- feeding was also viewed as more convenient. It is surprising
strating a significant difference by race. Females were also that many students demonstrated additional support for
more likely to know someone who breastfed and have higher breastfeeding by disagreeing with statements that women
154 Journal of Human Lactation 33(1)

Table 4.  Mean Comparisons in Attitude Scores Table 6.  Analysis of Effects for Breastfeeding Intentions

Variable M (SD) t-test p Variable Wald χ2 p ORa [95% CI]


Race 5.59 < .001 Race 1.954 .162 1.298 [0.900, 1.972]
  African Americana 57 (7.21)   Gender 1.043 .307 1.213 [0.837, 1.758]
 Caucasiana 61 (7.42)   Attitudes 58.171 < .001 1.803 [1.549, 2.098]
Gender 4.90 < .001 Breastfed as infant 43.907 < .001 4.567 [2.914, 7.157]
 Female 60 (7.97)   Know someone 7.797 .005 2.941 [1.379, 6.272]
 Male 58 (6.67)   who breastfed
Race within gender
Note. N = 696. CI = confidence interval; OR = odds ratio.
 Female 3.83 < .001 a
The OR for race = odds of yes for Caucasian, gender = odds of yes for
  African American 59 (7.72)   female, and breastfeeding exposure = odds of yes.
  Caucasian 62 (7.92)  
 Male 4.29 < .001
Beliefs about breastfeeding that correlate with breastfeed-
  African American 56 (6.41)  
ing intentions are formed prior to pregnancy. Thus, it is essen-
  Caucasian 59 (6.62)  
tial to start breastfeeding education and promotion before
Breastfeeding exposure
  Breastfed as an infant 61 (6.91) 8.04 < .001
women become pregnant. Including breastfeeding education
  Not breastfed as an infant 56 (7.66)   as part of the curriculum in high school health classes and col-
  Know someone who 59 (7.39) 3.69 < .001 lege courses could expose teens and young adults to breast-
breastfed feeding. Presentations about breastfeeding during health fairs
  Did not know someone 55 (7.53)   in the African American community could also expose the
who breastfed general population, encouraging acceptance of breastfeeding.
For some African Americans, this may be the only exposure to
Note. N = 696.
a
Cronbach’s alpha by race: African American = .70; Caucasian = .75.
breastfeeding due to higher rates of formula feeding in their
community. It is also essential for breastfeeding education to
target both women and men and be framed as the natural, first
Table 5.  Associations With High Probability of Breastfeeding
Intentions option instead of an alternative to formula feeding. In addition,
breastfeeding recommendations and education should occur
Variable n (%) χ2 p early in pregnancy during clinic visits and prenatal classes.
Race 19.95 < .001 Waiting to discuss breastfeeding until the third trimester or
  African American 175 (50)   upon hospital admission for delivery may be too late in the
 Caucasian 234 (67)   decision process to have an effect on initiation. At the hospital,
Gender 7.68 .006 women are often focused on the birth and welcoming the
 Female 223 (64)   infant. An attempt to promote breastfeeding only in the hospi-
 Male 186 (53)   tal may compete with the excitement of birth. If there is no
Breastfed as infant 299 (77) 124.93 < .001 intention to breastfeed before pregnancy or in the prenatal
Know someone 396 (62) 30.19 < .001 period, it may be unlikely that a woman will initiate breast-
who breastfed feeding. Therefore, every effort should be made to improve
breastfeeding intentions both prior to pregnancy and during
Note. N = 696.
the prenatal period. Any efforts to improve breastfeeding
exposure and attitudes among women and men for the popula-
should not breastfeed in public and fathers will feel left out if tion with the lowest breastfeeding rates could potentially
a woman breastfeeds. Beliefs about breastfeeding in public improve public acceptance of breastfeeding, increase the
have varied among studies. Although some students feel that probability of breastfeeding intentions, and reduce the racial
women should not breastfeed in public (Jefferson, 2014b; disparity in breastfeeding.
Kavanagh et al., 2012; Spear, 2006), the majority of health
science students were comfortable with public breastfeeding
(Dodgson et al., 2014). In this study, comfort with public
Limitations
breastfeeding could be due to the majority of the students Several limitations for this study were noted. First, the cross-
knowing someone who breastfed, which might give them sectional design with nonpregnant women prevented predic-
greater comfort with breastfeeding. In addition, college stu- tions about breastfeeding initiation and ability to conclude a
dents may experience exposure to breastfeeding education in causal relationship among breastfeeding exposure, attitudes,
nutrition, ethics, or health courses, and national breastfeeding intentions, and initiation. Second, breastfeeding intentions
campaigns could have made a good impression on young were measured at one point among a sample without children
adults to improve acceptance of breastfeeding in public. and may not reflect breastfeeding behavior later in life.
Jefferson 155

Table 7.  Analysis of Effects for Breastfeeding Exposure and Attitudes

Know someone who breastfeda Breastfed as infantb Attitude scorec

Variable Wald χ2 p ORd [95% CI] Wald χ2 p ORd [95% CI] βe [95% CI] p
Race 10.60 .001 2.72 [1.49, 4.98] 23.88 < .001 2.59 [1.77, 3.79] 3.10 [2.03, 4.17] < .001
Gender  7.56 .006 2.28 [1.28, 4.09]  2.26  .13 0.75 [0.51, 1.09] 2.73 [1.66, 3.80] < .001

Note. N = 696. CI = confidence interval; OR = odds ratio.


a
Nagelkerke R2 = .065. bNagelkerke R2 = .068. cR2 = .077. dThe OR for race = odds of yes for Caucasian and gender = odds of yes for female. eCoefficient
gives the estimated increase in mean for Caucasian race and female gender.

Intentions can vary as time passes and unforeseen events through 6 months by race. Research should include infant-
occur. Third, breastfeeding exposure was identified only by feeding exposure by close contact with a breastfeeding
being breastfed as an infant and knowing someone who has woman, academic education, and general media sources.
breastfed, which allowed only a general measure of breast- Assessing the frequency or duration of infant-feeding expo-
feeding exposure. Students were not asked to report exposure sure would provide an in-depth view of exposure and asso-
to formula feeding, which prevented evaluation of the associa- ciation with breastfeeding outcomes. This evidence could be
tion with infant-feeding attitudes and intentions. Fourth, the useful in the development of interventions tailored to
Cronbach’s alpha of the IIFAS for the total sample was .72, .70 improve breastfeeding exposure among African American
for African Americans (mean item score = 3), and .75 for women.
Caucasians (mean item score = 4). Although greater than or
equal to .80 is desired for internal consistency, greater than or
Conclusion
equal to .70 is acceptable for group-level comparisons (Polit &
Beck, 2004). The reliability coefficient in this study may have External factors, such as exposure and attitudes, have a
resulted from the statistically significant difference in mean stronger association with breastfeeding intentions. However,
scores by race and gender or due to the number of neutral differences in race and gender predicted exposure and atti-
responses for some items, indicating that some participants tudes, which may contribute an indirect effect on breastfeed-
may have been unsure about the infant-feeding statements. ing intentions. Interventions are needed for vulnerable
Last, data were collected from one college in the Midwest, populations to increase breastfeeding exposure and improve
which limits the generalizability of findings to other regions. attitudes about breastfeeding prior to pregnancy and during
Nonetheless, findings highlight the association of breastfeed- early pregnancy, which may affect intentions. Currently,
ing exposure and attitudes with breastfeeding intentions. breastfeeding interventions targeting minorities typically
focus on prenatal education, initiation at birth, and support in
the postpartum period (Chapman & Perez-Escamilla, 2012).
Implications for Research Interventions focused on improving breastfeeding intentions
Reasons for the persistent racial disparity in breastfeeding are limited. Including men in breastfeeding interventions is
remain a complex and important phenomenon. Findings of also essential in order to improve paternal support for breast-
this study indicate that breastfeeding exposure and attitudes feeding as the best option for infants. This needed evidence
may be one explanation for the variability in the racial dis- may help close the gap in the racial disparity in breastfeeding
parity in breastfeeding, as reflected by Caucasian race pre- and reach the Healthy People 2020 objectives.
senting as the strongest predictor for breastfeeding exposure
and attitudes. Because breastfeeding exposure and attitudes Acknowledgments
significantly contributed to breastfeeding intentions, it is The author would like to thank Richard Madsen, PhD, for assis-
important for research to focus on identifying ways to tance with statistical analyses.
increase exposure and attitudes for African Americans. Many
African Americans may have less personal contact with Declaration of Conflicting Interests
breastfeeding women due to the low rates of breastfeeding The author declared no potential conflicts of interest with respect to
and less opportunity for any exposure in courses if not attend- the research, authorship, and/or publication of this article.
ing college. Inclusion of adequate samples from various
racial groups would allow comparisons by race and ability to Funding
explore reasons for the racial disparity in breastfeeding. The author disclosed receipt of the following financial support for
Longitudinal studies could examine the correlation of exter- the research, authorship, and/or publication of this article: This
nal factors with prenatal breastfeeding intentions and the study was funded by the University of Missouri–Columbia Richard
ability of intentions to predict actual breastfeeding initiation Wallace Faculty Incentive Award.
156 Journal of Human Lactation 33(1)

References Jefferson, U. T. (2015). Predictors of breastfeeding attitudes among


college-educated African Americans. Research and Theory for
Centers for Disease Control and Prevention. (2010). Racial and
Nursing Practice: An International Journal, 29(3), 189–199.
ethnic differences in breastfeeding initiation and duration,
Jones, K. M., Power, M. L., Queenan, J. T., & Schulkin, J. (2015).
by state—National Immunization Survey, United States,
Racial and ethnic disparities in breastfeeding. Breastfeeding
2004-2008. Morbidity and Mortality Weekly Report, 59(11),
Medicine, 10, 186–196. doi:10.1089/bfm.2014.0152
327–334.
Kavanagh, K. F., Lou, Z., Nicklas, J. C., Habibi, M. F., & Murphy, L.
Centers for Disease Control and Prevention. (2016). Breastfeeding
T. (2012). Breastfeeding knowledge, attitudes, prior exposure,
among U.S. children born 2002-2013, CDC national immuni-
and intent among undergraduate students. Journal of Human
zation survey. Retrieved from http://www.cdc.gov/breastfeed-
Lactation, 28, 556–564. doi:10.1177/0890334412446798
ing/data/NIS_data/index.htm
Marrone, S., Vogeltanz-Holm, N., & Holm, J. (2008). Attitudes,
Chambers, J. A., McInnes, R. J., Hoddinott, P., & Alder, E. M.
knowledge, and intentions related to breastfeeding among
(2007). A systematic review of measures assessing mothers’
university undergraduate women and men. Journal of Human
knowledge, attitudes, confidence and satisfaction towards
Lactation, 24, 186–192. doi:10.1177/0890334408316072
breastfeeding. Breastfeeding Review, 15(3), 17–25.
Nommsen-Rivers, L. A., Chantry, C. J., Cohen, R. J., & Dewey,
Chapman, D. J., & Perez-Escamilla, R. (2012). Breastfeeding
K. G. (2010). Comfort with the idea of formula feeding helps
among minority women: Moving from risk factors to interven-
explain ethnic disparity in breastfeeding intentions among
tions. Advances in Nutrition: An International Review Journal,
expectant first-time mothers. Breastfeeding Medicine, 5(1),
3(3), 95–104.
25–33. doi:10.1089/bfm.2009.0052
de la Mora, A., Russell, D., Dungy, C., Losch, M., & Dusdieker, L.
Polit, D., & Beck, C. (2004). Nursing research: Principles and
(1999). The Iowa Infant Feeding Attitude Scale: Analysis of
methods (7th ed.). Philadelphia, PA: Lippincott Williams &
reliability and validity. Journal of Applied Social Psychology,
Wilkins.
29, 2362–2380.
Shaker, I., Scott, J. A., & Reid, M. (2004). Infant feeding attitudes
Dodgson, J. E., Bloomfield, M., & Choi, M. (2014). Are health sci-
of expectant parents: Breastfeeding and formula feeding.
ence students’ beliefs about infant nutrition evidence-based?
Journal of Advanced Nursing, 45, 260–268.
Nurse Education Today, 34, 92–99.
Eidelman, A. I., Schanler, R. J., Johnston, M., Landers, S., Noble, Seidel, A. K., Schetzina, K. E., Freeman, S. C., Coulter, M. M., &
L., Szucs, K., & Viehmann, L. (2012). Breastfeeding and the Colgrove, N. J. (2013). Comparison of breast-feeding knowledge,
use of human milk. Pediatrics, 129, e827–e841. doi:10.1542/ attitudes, and beliefs before and after educational intervention
peds.2011-3552 for rural Appalachian high school students. Southern Medical
Froehlich, J., Boivin, M., Rice, D., McGraw, K., Munson, E., Journal, 106, 224–229. doi:10.1097/SMJ.0b013e3182882b8f
Walter, K. C., & Bloch, M. K. (2013). Influencing univer- Sittlington, J., Stewart-Knox, B., Wright, M., Bradbury, I., & Scott,
sity students’ knowledge and attitudes toward breastfeeding. J. A. (2007). Infant-feeding attitudes of expectant mothers in
Journal of Nutrition Education and Behavior, 45(3), 282–284. Northern Ireland. Health Education Research, 22, 561– 570.
Ip, S., Chung, M., Raman, G., Chew, P., Magula, N., DeVine, D., Spear, H. J. (2006). Baccalaureate nursing students’ breastfeeding
. . . Lau, J. (2007). Breastfeeding and maternal and infant health knowledge: A descriptive survey. Nurse Education Today, 26,
outcomes in developed countries. Evidence Report/Technology 332–337.
Assessment, (153), 1–186. Street, D. J., & Lewallen, L. P. (2013). The influence of culture
Jefferson, U. T. (2014a). A review of the literature regarding deter- on breast-feeding decisions by African American and white
minants of breastfeeding among black women in the United women. Journal of Perinatal & Neonatal Nursing, 27(1), 43-
States. The Journal of the National Black Nurses Association, 51. doi:10.1097/JPN.0b013e31827e57e7
25(2), 55–63. Vari, P., Vogeltanz-Holm, N., Olsen, G., Anderson, C., Holm, J.,
Jefferson, U. T. (2014b). Infant feeding attitudes and breastfeeding Peterson, H., & Henly, S. (2013). Community breastfeeding
intentions of black college students. Western Journal of Nursing attitudes and beliefs. Health Care for Women International, 34,
Research, 36, 1338–1356. doi:10.1177/0193945913514638 592–606.

You might also like