Extended Maternity Leave and Breastfeeding in Active Duty Mothers

We conducted a retrospective cohort study using data collected from DoD electronic health care records (EHRs) of all infants born to active duty mothers who were delivered at BAMC during the calendar years 2014 and 2016. EHR platforms were the Armed Forces Health Longitudinal Technology Application for outpatient records and Essentris for inpatient records. A single military treatment facility (MTF) was used because of the study design and the availability of delivery records. A control group could not be included because of the retrospective study design. The year 2015 was excluded because the Navy’s policy changed during the year (maternity leave was temporarily increased to 18 weeks). Therefore, 2015 was excluded from this analysis.¹⁹ 

This study was reviewed and approved by the Institutional Review Board at BAMC (Fort Sam Houston, TX). No personally identifiable data were collected, and all data were stored on secured government computers.

No major changes related to breastfeeding support or breastfeeding policy occurred at BAMC during the study period. BAMC achieved and maintained Texas Ten Step designation for breastfeeding before the study period. Texas Ten Step is a designation used to recognize hospitals that have implemented policies aligned with the World Health Organization Ten Steps to Successful Breastfeeding.²⁰ 

Data were obtained from the inpatient record and from outpatient routine well-child visits by using the DoD EHR systems. The inpatient EHR was used to determine maternal and infant demographic data as well as documented feeding at time of hospital discharge. Demographic data were evaluated by using information that was reported in the literature to have an effect on breastfeeding success (eg, delivery method and maternal parity) (Table 1).^9,–13 The outpatient EHR was used to determine the type of feeding at the aforementioned well-child checks. Feeding type was determined on the basis of documentation in the EHR at each time point. At each visit, parents are asked about infant nutrition (breast milk versus formula) so that the infant’s nutritional status can be assessed and the appropriate anticipatory guidance provided. The Armed Forces Health Longitudinal Technology Application has several provider templates wherein feeding method can be easily documented. Although typically well documented, in the absence of explicit documentation, any ambiguous data were recorded as “no breast milk.” Parent questionnaires were not used at any point to determine feeding method or intentions.

All infants born to active duty military mothers during the defined time period at BAMC were included for evaluation. Infants were excluded if they did not receive well-child care within an MTF until at least 2 months of age. Examples of this exclusion criterion include infants who received care in civilian health care systems, infants who did not attend regularly scheduled well-child visits, or infants whose parents separated from the military. Infants were also excluded if they had a documented medical contraindication to breast milk or died before 12 months of age. Multiple-gestation deliveries were counted as an independent metric (singleton versus multiple delivery) during evaluation of the study demographics. However, only a singular mother-infant dyad was analyzed to maintain equal numbers of infants and mothers in the study. For every multiple delivery, the data of the youngest living infant were used.

All data were de-identified during collection; therefore, mothers who delivered in both study years were included without any augmentation to the analysis. No mothers had 2 discrete pregnancy and delivery events in the same study year.

The primary outcome of breastfeeding was defined as the provision of any maternal breast milk. Exclusive breastfeeding is defined by The Joint Commission as “a newborn receiving only breast milk and no other liquids or solids except for…medicines.”²¹ We did not evaluate complementary food consumption or how it affected breast milk use. As such, when we refer to exclusive breastfeeding, we indicate breast milk consumption without any provision of formula. Secondary outcomes included differences between military-branch affiliation (eg, Army) and military rank, defined as officer versus enlisted.

A statistical analysis was performed by using JMP 13.0 (SAS Institute, Inc, Cary, NC). Comparisons of demographic data between the groups were evaluated by using a Wilcoxon rank test. Categorical variables were analyzed by using a χ² analysis for each time point. Continuous variables were analyzed by using a t test. A survival analysis with Kaplan-Meier plots and the log-rank test was used to evaluate the primary outcome of increased duration. Finally, a proportional hazard regression was used to evaluate differences in the time to cessation of breastfeeding between the study periods after we adjusted for confounding variables such as previous breastfeeding success, level of partner support, and cultural influences on breastfeeding. Hazard ratios (HRs) and their corresponding 95% confidence intervals were determined for these results.²² Because of loss to follow-up throughout the year (eg, separation from the military), each data point was analyzed by using only the number of infants still enrolled at any given point.

The charts of all active duty mothers who delivered during 2014 and 2016 were reviewed. Of the 952 infants reviewed, 3 were excluded medically, and 66 were excluded because of loss to follow-up in an MTF. There were 24 twin and triplet deliveries, which accounted for 26 excluded A twins and A and B triplets. After exclusion, 423 dyads in 2014 and 434 dyads in 2016 were included for statistical evaluation. There was ∼10% loss of participants by the 12-month visit in each group (Table 2). The study years 2014 and 2016 had similar demographic compositions (Table 1). Military demographics, including rank and branch, were also equally distributed. Similar to the general composition of the military, the bulk of the personnel were enlisted. The primary branch represented in this study was the Air Force (AF), followed by the Army. There were relatively few births among Navy, Marine, or Coast Guard personnel. A t test revealed no differences in maternal age, gestational age, or BMI. We noted a significant increase in NICU admissions in 2016 (20.7%; up from 12.1% [P = .003]).

There was no overall change in breastfeeding initiation (95.3% vs 95.2%; P = .63) between 2014 and 2016. Documented establishment and duration of any breastfeeding increased at the 2- (72.4% vs 81.5%; P = .002), 4- (56.7% vs 72.2%; P < .001), 6- (46.5% vs 60.3%; P < .001), and 9-month (34.9% vs 45.4%; P = .003) visits in the 12-week leave cohort. There was no significant difference noted at the 12-month visit (24.8% vs 29.3%; P = .16). Exclusive breastfeeding also increased at the 2- (47.2% vs 56.4%; P = .007), 4- (36.4% vs 47.5%; P = .001), and 6-month (29.3% vs 37.3%; P = .016) visits. There was a trend toward improvement at the 9-month (20.6% vs 26.4%; P = .056) visit (Table 3).

No individual branch of the military had significantly higher rates of breastfeeding initiation or duration. Similarly, there was no significant difference in breastfeeding initiation between ranks. Officer rank, however, was associated with higher breast milk use at all subsequent time points (Fig 1). There was an increase in breastfeeding for the officer cohort between the 2- and 9-month data points; however, it did not achieve statistical significance. Enlisted personnel, however, increased breastfeeding at the 2- (66.2% vs 78.5%; P < .001), 4- (48.6% vs 64.4%; P < .001), 6- (38.5% vs 53.7%; P < .001), and 9-month (30.1% vs 39.3%; P = .023) visits. There was a trending increase (18.8% vs 25%; P = .08) at 12 months (Table 3).

As discussed previously, the Navy policy was in transition during most of 2016. The AF and Army had 6 vs 12 weeks of leave in 2014 and 2016, respectively, whereas the Navy had 6 vs ≥12 in 2014 and 2016, respectively. A sensitivity analysis was performed on the data with and without Navy inclusion, with no difference in outcomes for either study year.

The Kaplan-Meier analysis revealed a highly significant difference in time to cessation of breastfeeding for marital status and rank status (log-rank P < .001) as well as between 2014 and 2016 (log-rank P = .006). We did not find any significance for maternal race, maternal age, infant sex, parity, multiple births, or mode of delivery with respect to cessation of breastfeeding. On the basis of these results, a multivariate proportional hazard regression was performed, with adjustment for marital status, rank status, gestational age, and year. Only year (HR = 1.23; P = .016) and rank (HR = 1.65; P < .001) remained significant in the multivariate proportional hazard regression model (Table 4).

In 2016, there was a significant increase in breastfeeding duration and exclusive breastfeeding in the active duty population cared for at our facility. Although unchanged, the breastfeeding initiation rate of 95% in the military remained well above the CDC-reported national average of 83.2% of infants ever breastfed.⁵ Breastfeeding and breastfeeding exclusivity increased at all measured time points, surpassing several CDC Healthy People 2020 goals.⁴ 

The study design does not allow for the conclusion of direct causation, and it is important to note that some of the results could be related to the nationwide increase in breastfeeding success during this same time period. The CDC reports that infants receiving any breast milk and infants receiving breast milk exclusively at 6 months increased by 2.4% and 3.5%, respectively, from 2014 to 2016.⁴ At our facility, the same metrics increased by 13.8% and 8%, respectively, during the same time period.

Our data are consistent with data from available literature that reveal that longer maternity leave and delayed return to work correlate with longer breastfeeding duration.^12,–15 Authors of previous studies have identified that demographic factors, such as socioeconomic status, have significant impact on establishing and maintaining breastfeeding. In general, officers must have at least a college degree, whereas it is not required for enlisted personnel. This is important because authors of several studies cite advanced education as a factor associated with higher rates of breastfeeding success.¹⁰ In the military setting, rank can be used as a surrogate for educational level, income, and age. Although officer and enlisted groups improved their breastfeeding duration with longer maternity leave, a larger increase was demonstrated in the enlisted population for both breastfeeding duration and exclusive breastfeeding. This may signify potentially more robust effects in this particular demographic, a finding that is consistent with civilian data.^10,–12 

There were significantly more NICU admissions in 2016. There was no concurrent change in policy to account for the higher rate of admission. It is interesting to note that despite the higher number of admissions, infants admitted to the NICU had better breastfeeding success in 2016 vs 2014. Although there is literature to support a positive association between NICU admission and breastfeeding in preterm infants, the opposite is true for term infants. Our results are contrary to those of the study by Colaizy and Morriss²³ in that we had increased breast milk use in term infants admitted to the NICU.

A primary limitation of this study is that it is a retrospective study that relied on a single MTF and accurate documentation of feeding methods in outpatient records. In either study year, it is possible that a service member could have extended their maternity leave using regular leave (vacation time). These data are not tracked in the EHR, so it is impossible to know if a service member chose to do so. The demographics are even both years, so this effect should be negligible and should not impact final outcome data. Breastfeeding success in previous pregnancies is not uniformly documented in the EHR and therefore could not be evaluated. Likewise, social support at home outside of marital status could not be reliably assessed, both of which represent potential confounding variables to the data.

Additional limitations include increased awareness of insurance coverage of home breast pumps and supplies, access to outpatient lactation support, and increased employer provision of appropriate breastfeeding and pumping locations in the military setting. Breast pumps became covered by Tricare, the military insurance provider, in late 2014. The majority of women who delivered in 2014 either had to buy a pump or did not otherwise have access. However, despite the assumption that access to a breast pump would support breastfeeding success, there is no definitive literature to support this claim.²⁴ Furthermore, we could not control for national trends occurring outside the hospital, such as increasing awareness and encouragement of breastfeeding, which have led to a nationwide increase in breastfeeding rates.⁴ 

Although there were no significant changes noted among the military branches evaluated, it should be noted that the Army and AF were primarily represented. Navy personnel at the study site accounted for only 6.6% and 5.3% of personnel each year. Army and AF predominance in the study is secondary to the specific MTF and the geographic region reviewed. A similar study in which authors look at MTFs, with a primary focus on the Navy or Marine Corps, could be warranted. Additionally, the base at which this study was conducted and the majority of the bases supported by large MTFs are mainly training commands. These bases support readiness and education but are not typically primarily responsible for supporting large-scale combat deployment. Evaluating other military branches, geographic regions, varying operations and support objectives could be of use to the military to identify if these trends are maintained in other environments.

Differences in health care use for mothers and infants during the study period were not evaluated. However, authors of future studies could evaluate these differences (ie, the number of infant sick visits for ear infections and asthma) to determine if better breastfeeding success correlates with lower health care use, a concept repeatedly borne out in the literature and supported by the American Academy of Pediatrics.² 

Finally, the study was not powered to detect statistical significance among the officer group. A post hoc analysis suggested that in a doubled sample size, the officer results would reach significance.

There has been a statistically significant increase in breastfeeding duration through 9 months and breastfeeding exclusivity through 6 months in active duty mothers since the change to a 12-week policy despite no change in breastfeeding initiation. Mothers were able to provide breast milk significantly longer in the 2016 cohort than in the 2014 cohort, surpassing the goals established by the CDC.

In addition, officers continue to have higher initiation and duration than their enlisted counterparts, although the change in policy seems to have had a stronger impact on overall breastfeeding success in the enlisted community. The implications of this specific result suggest that extending maternity leave has benefit overall but more pronounced benefit in groups that are less likely to have breastfeeding success. These findings can be easily interpreted as generalizable to the workforce overall.

Finally, these data affirm the DoD’s decision to allow longer maternity leave as it relates to breastfeeding. It also supports further studies to assess if this holds true at other locations in the DoD. Ultimately, the highest affirmation of benefit for the military would be establishing if higher breastfeeding success translates into improved job satisfaction and military retention, as stated as an intention by Defense Secretary Ash Carter.¹⁹ 

We thank Dr Kaashif Ahmad for his guidance and direction as well as for his thoughtful insights in revising the article and Dr Jazmin Lesnick for her help in data collection.

AF

Air Force

BAMC

Brooke Army Medical Center

CDC

Centers for Disease Control and Prevention

DoD

Department of Defense

EHR

electronic health care record

HR

hazard ratio

MTF

military treatment facility