- Click here for explanation of classification scheme.

To determine whether an individualized, prenatal and post-natal lactation consultant intervention resulted in increased cumulative intensity of breastfeeding up to 52 weeks.

• English- or Spanish-speaking
• Twin or singleton pregnancy (originally, but during data analysis, only mothers of singletons were included)
• Intent to keep the infant
• Plans to remain in the center and its affiliated hospital system for prenatal and infant care up to 12 months
• At least two contact telephone numbers
• Gestation before 24 weeks.

• HIV-positive status
• Chronic therapy with medications incompatible with breastfeeding
• Pre-gestational diabetes mellitus
• Human T-cell leukemia virus-1
• Breast reduction surgery
• Hepatitis B or C.

Recruitment

Recruited from two hospital-affiliated health centers:

• From one center, participants were recruited by a research assistant from the center's mandatory prenatal class
• From the other center, social work staff members recruited participants.

Design

• Eligible participants randomly assigned to:
  • Intervention group that received lactation consultant intervention
  • Control group that received standard care.
• Data collection included:
  • One prenatal baseline interview
  • Eight telephone interviews at one, two, three, four, six, eight, 10 and 12 months after birth.

Blinding Used

Research assistant and lactation consultants were not blinded to treatment group, but data were not shared between research assistants and lactation consultants.

Intervention

Intervention included:

• Two attempted individual meetings with a lactation consultant, prenatally and post-partum in the hospital or at home  
• Telephone consultation up to 12 months
• Breastfeeding education and counseling
• Assistance garnering support from families, schools, workplaces and health care provider
• Nursing bra
• Manual or mini-electric breast pumps, depending on circumstance.

Statistical Analysis

• Unadjusted differences between groups tested for significance with T-tests, chi-square tests and the Mantel-Haenszel test
• Differences between sites on 60 baseline variables were analyzed
• Bivariate associations between breastfeeding intensity scores and other variables analyzed with Wilcoxon rank sum tests and Spearman rank correlations
• Logistic regression models analyzed breastfeeding intensity scores as outcome variables with significant covariates, treatment group, interaction between treatment group and country of origin as independent factors
• Analyses used a intention-to-treat model.

Timing of Measurements

Data collection included:

• One prenatal baseline interview
• Eight telephone interviews at one, two, three, four, eight, 10 and 12 months after birth.

Dependent Variables

• Breastfeeding intensity through Week 13 (total, 13)
• Breastfeeding intensity through Week 52 (total, 52)
  • Weekly scores based on Index of Breastfeeding Status that measures the percentage of breast milk an infant receives, compared with the total amount of feedings. A score of one is exclusive breast milk, whereas a score of seven is exclusive formula.
  • Intensity scores derived by summing intensity scores obtained each week
  • Dichotomous variable created (low vs. high breastfeeding) by using the median as a cut-point.

Independent Variables

•  Demographics
  • Age
  • High school graduate
  • Marital status
  • Birthplace (foreign born)
  • Race or ethnicity
  • Medicaid recipient
  • Parity
  • Prior experience breastfeeding
  • Intentions to breastfeed.
• Birth and postpartum experiences
  • Birth location (study hospital)
  • Delivery time (day or night)
  • Who baby left hospital with (mother)
  • Birth type (vaginal or Cesarean)
  • Birth weight (low birth weight or normal weight)
  • Neonatal Intensive Care Unit (NICU) (yes or no)
  • Roomed in (yes or no)
  • First breastfeeding attempt (zero to four hours; after four hours)
  • Advised on breastfeeding by prenatal care provider (yes or no)
  • Asked if wanted to nurse in delivery room by hospital staff (yes or no)
  • When first considered how to feed infant (before pregnancy, first trimester or later).
• Treatment group
  • Intervention
  • Control.

Initial N

382 women were randomly assigned:

• 188 in Intervention Group
• 194 in Control Group.

Attrition (Final N)

304 women with at least one post-natal follow-up data point.

• 145 in Intervention Group
• 159 in Control Group.

Age

Mean age in years ± standard deviation (SD)

• Intervention Group: 25.48±6.39
• Control Group: 24.99±6.09.

Ethnicity (N)

• Black
  • 52 (35.9%) in Intervention Group
  • 58 (36.5%) in Control Group.
• Hispanic
  • 82 (56.6%) in Intervention Group
  • 90 (56.6%) in Control Group.

Other Relevant Demographics

• No statistically significant difference between groups in age, race or ethnicity, high school education, marital status, foreign birth, Medicaid coverage, parity, previous breastfeeding experience or intentions with regard to breastfeeding
• No statistically significant difference between groups in birth and postpartum experience variables, including infant birth weight (low birth weight vs. normal weight).

Location

Bronx, New York.

Key Findings

• Any breastfeeding: The Intervention Group was significantly more likely to breastfeed at each week up to and including Week 20 (53.0% vs. 39.3%, P<0.028) with the exception of Week 18
• Breastfeeding of 50% or more: Significantly more women in the Intervention Group gave 50% or more breast milk up to and including Week Nine (45.8% vs. 33.1%, P<0.030)
• Exclusive breastfeeding:  Exclusive breastfeeding rates were not significantly different between intervention and control groups at any of the 52 weeks observed.

Logistic Regression: Risk of Low Breastfeeding (Compared to High Breastfeeding) (N=304)

Variables		Week 13 Odds Ratio [95% Confidence Interval (CI)]	Week 52 Odds Ratio [95% Confidence Interval (CI)]
Intent at Baseline	Undecided	3.04* (1.22 to 7.58)	3.02* (1.21 to 7.56)
	Formula	10.51* (2.97 to 37.15)	10.05* (3.04 to 33.20)
	Combined	1.57 (0.85 to 2.87)	2.22* (1.19 to 4.15)
	Breast milk (referent)	1.00	1.00
Logarithm of Age	Years	0.92* (0.88-0.96)	N/A
Age	Less than 24 years	N/A	3.15* (1.83 to 5.42)
	24 years or more (referent)	N/A	1.00
Race or Ethnicity	Black	0.44* (0.24 to 0.79)	0.37* (0.20 to 0.67)
	Other	1.75 (0.60 to 5.13)	1.40 (0.48 to 4.10)
	Hispanic (referent)	1.00	1.00
Medicaid	No	1.05 (0.61 to 1.82)	0.71 (0.41 to 1.23)
	Yes (referent)	1.00	1.00
Treatment Group/ Country of Origin	Control/US-born	5.22* (2.43 to 11.22)	5.25* (2.44 to 11.29)
	Control/foreign-born	0.90 (0.40 to 2.04)	1.06 (0.47 to 2.40)
	Intervention/US-born	1.67 (0.78 to 3.53)	1.15 (0.54 to 2.45)
	Intervention/foreign-born (referent)	1.00	1.00
Model R2		0.24	0.25

* P<0.05
N/A: Data not given in paper's table.

• In bivariate analysis, variables with associations yielding P-values of less than 0.20 were used in initial multivariate models
• Women who were undecided about feeding intentions and those intending to formula feed exclusively at baseline had a significantly greater risk of low breastfeeding at both 13 and 52 weeks. At 52 weeks only, women who planned to both breast and formula feed combined had a significantly greater risk of low breastfeeding. 
• Younger women were at greater risk of low breastfeeding at both 13 and 52 weeks
• Black women were at a significantly lower risk of low breastfeeding compared to the Hispanic referent group at both 13 and 52 weeks
• Women who were in the Control Group and born in the U.S. had a significantly greater risk of low breastfeeding at both 13 and 52 weeks, compared to foreign-born women in the intervention group
• Logistic regression models also analyzed a sub-sample of the participants who had at least one other child at baseline. These models yielded similar results, except being undecided about feeding intentions no longer a significant risk factor for low breastfeeding at both 13 and 52 weeks and planning to combine breast and formula feeding was no longer a significant risk factor at 52 weeks. Women who had not breastfed previously had a significantly greater risk of low breastfeeding at 52 weeks but not at 13 weeks, compared to women who had previously breastfed.
• Alternative logistic regression models used the same set of covariates, except that treatment group was used in place of the treatment group or country of origin interaction terms. ORs and 95% CIs for covariates in the alternative logistic regression model were essentially unchanged from those shown above.

• At three and 12 months, control subjects receiving standard care in urban primary care health centers had 90% and 150% increased risk of low breastfeeding, respectively, compared with the Intervention Group. We observed similar rates of exclusive breastfeeding between the two groups, which suggests that the intervention was not effective in helping women to breastfeed exclusively. Our most striking finding was the strength with which country of origin modified the effect of treatment group. US-born women were five to eight times as likely to have low breastfeeding intensity, depending on observation period and parity, compared with the foreign-born Control, foreign-born Intervention and US-born Intervention Groups. Therefore, our intervention was highly effective in raising the breastfeeding level of US-born women to that of foreign-born women, whose scores did not differ according to treatment group. 
• In contrast to the aforementioned results, race and ethnicity did not modify the effect of treatment group significantly. That is, the intervention was no more effective among one racial or ethnic group than another. Furthermore, in sharp contrast to recent national studies, black women were 50% to 60% more likely to have a high intensity of breastfeeding, compared with a reference group of women who identified themselves as Hispanic. Presumably, this finding is related to demographic distributions in the study's locale, i.e., New York City. New York City experienced a 41% increase in the West Indian population between 1990 and 2000, nearly all of whom identify themselves as black. West Indians represented 25% of the black population in New York City in 2000. Compared with the United States, West Indian nations have considerably higher rates of any and exclusive breastfeeding and longer duration of breastfeeding.
• This "best-practices" intervention was effective in increasing breastfeeding duration and intensity. Breastfeeding promotion should focus on US-born women and exclusive breastfeeding.

Government:

United States Department of Agriculture, the Maternal and Child Health Bureau, and the Agency for Healthcare Quality and Research