An Online Calculator to Estimate the Impact of Changes in Breastfeeding Rates on Population Health and Costs

Abstract

Objective:

We sought to determine the impact of changes in breastfeeding rates on population health.

Materials and Methods:

We used a Monte Carlo simulation model to estimate the population-level changes in disease burden associated with marginal changes in rates of any breastfeeding at each month from birth to 12 months of life, and in rates of exclusive breastfeeding from birth to 6 months of life. We used these marginal estimates to construct an interactive online calculator (available at www.usbreastfeeding.org/saving-calc). The Institutional Review Board of the Cambridge Health Alliance exempted the study.

Results:

Using our interactive online calculator, we found that a 5% point increase in breastfeeding rates was associated with statistically significant differences in child infectious morbidity for the U.S. population, including otitis media (101,952 cases, 95% confidence interval [CI] 77,929–131,894 cases) and gastrointestinal infection (236,073 cases, 95% CI 190,643–290,278 cases). Associated medical cost differences were $31,784,763 (95% CI $24,295,235–$41,119,548) for otitis media and $12,588,848 ($10,166,203–$15,479,352) for gastrointestinal infection. The state-level impact of attaining Healthy People 2020 goals varied by population size and current breastfeeding rates.

Conclusion:

Modest increases in breastfeeding rates substantially impact healthcare costs in the first year of life.

Introduction

Breastfeeding is associated with important health outcomes for mothers and children across the lifespan. To the extent that these associations are causal, interventions to enable women to initiate and sustain breastfeeding will reduce morbidity and mortality and lower healthcare costs, particularly in light of historically low breastfeeding rates in the United States.¹ Understanding the population health impact of changes in breastfeeding rates can inform investments in policies that enable women to breastfeed.

Estimating the impact of changes in breastfeeding on health outcomes is complex, because associations vary for different health conditions, depending on the prevalence of the disease and the dose of breastfeeding associated with a change in disease risk. Our group previously used a Monte Carlo simulation model to estimate the population-level changes in disease burden associated with a change from current, 2012 breastfeeding rates to optimal breastfeeding.² Recognizing that ∼10% of mothers and infants are not able to breastfeed,³ optimal breastfeeding was defined as 90% of women exclusively breastfeeding each of her children for 6 months and continuing to breastfeed for 12 months, as recommended by major U.S. medical organizations.^4–6 We found that, across the lifetime of a cohort of women born in a single year, a change from current to optimal breastfeeding would prevent 2,619 maternal deaths (95% confidence interval [CI] 1,978–3,259) and 721 child deaths (543–899).² Given a steady-state population, these results approximate the annual impact of 90% optimal breastfeeding for the population as a whole.

Although 90% optimal breastfeeding is a worthy goal, existing interventions have produced more modest changes in breastfeeding rates.^7,8 We therefore sought to develop an approach to estimate the impact of smaller changes in breastfeeding rates, and to estimate impact within an individual state in the United States. Using the same Monte Carlo simulation model, we calculated the marginal effect of a 10% absolute change in any breastfeeding at each month from 0 to 12 months postpartum, and in exclusive breastfeeding at each month from 0 to 6 months postpartum. We used these data to create an online calculator, which allows users to estimate the impact of a change in breastfeeding rates on disease rates as well as medical costs and societal costs due to premature death in a population of women and their children.

Materials and Methods

The methods for our Monte Carlo simulation have previously been described in detail.² Briefly, we modeled health outcomes for a cohort of women who were aged 15 in 2002 and their children. Women were modeled through age 70, and their offspring were modeled through age 20. We used population birth rates for 2012 to model number of births for each modeled woman across her lifetime. We considered health outcomes for which there are consistent data supporting a protective association with breastfeeding. For children, we modeled acute lymphoblastic leukemia,⁹ acute otitis media,¹⁰ Crohn's disease and ulcerative colitis,^11–13 gastrointestinal infection,¹⁴ lower respiratory tract infection requiring hospitalization,¹⁰ necrotizing enterocolitis (NEC),¹⁵ sudden infant death syndrome,¹⁶ and, for non-Hispanic White (NHW) children, childhood obesity.¹⁷ For mothers, we modeled premenopausal ovarian cancer,¹⁸ breast cancer,¹⁹ diabetes,²⁰ hypertension,²¹ and myocardial infarction.²²

We sought to calculate the population health impact of a given change absolute rates of any breastfeeding from birth through 12 months, and in exclusive breastfeeding from birth through 6 months. We assumed that the health impact of a change in breastfeeding rates would be additive; for example, an absolute increase in exclusive breastfeeding rates at 6 months from 20% to 25%—a 5% absolute difference—would have half the impact of an absolute increase in exclusive breastfeeding rates at 6 months from 20% to 30%—a 10% absolute difference. To estimate the impact of an absolute difference in exclusive and/or total breastfeeding rates at a given month, we ran separate models, varying each by the breastfeeding rate at a single time point.

For example, in model A, we simulated a rate of any breastfeeding of 90% from 0 to 12 months postpartum, and a rate of exclusive breastfeeding of 80% from 0 to 6 months postpartum. In model B, breastfeeding rates were identical to model A, except that we simulated an exclusive breastfeeding rate at 6 months of 70%. Thus, the difference in disease burden between model A and model B approximated the effect of a 10% change in exclusive breastfeeding at 6 months, holding all else constant. To calculate the effect of a 10% change at each of the 20 time points, 0–12 months for any breastfeeding and 0–6 months for exclusive breastfeeding, we modeled outcomes under 21 different breastfeeding rates (Appendix Table A1).

For each of the 21 models, we performed 5,000 replications, modeling 100,000 women per replication. We chose a smaller number of replications than used in our previous work² to allow for greater uncertainty given that the calculator would be used to predict changes for smaller populations (e.g., individual states, communities). As associations between breastfeeding and disease outcomes in the literature are estimates, we drew key parameters at random from triangular distributions covering the range of estimates available in the literature associating breastfeeding with maternal and pediatric disease outcomes, centered on the point estimate provided in the literature and a distribution width related to the standard error or CI reported in the literature.

We analyzed variability in the diseases over this population; this analysis represents a form of probabilistic sensitivity analysis. To calculate CIs for maternal and childhood diseases, we randomly grouped the replications into batches of 20, representing 2 million women to approximate the 1.994 million women aged 15 in 2002, and then identified the 2.5% and 97.5% for each outcome.

We used the results of these 21 models to calculate the effect of a change in breastfeeding rates as described earlier. This process was repeated for 13 time points for any breastfeeding (month 0 through month 12) and for 7 time points for exclusive breastfeeding (month 0 through month 6). CIs for differences were added in quadrature, for example, the upper confidence interval (UCI) for Model B to Model A = (UCI_A² + UCI_B²)^1/2. Thus, for each disease outcome, we calculated a total of 20 marginal effect estimates, with upper and lower confidence limits (Appendix Table A2).

Using these marginal effects, we constructed a calculator to estimate the effect of changes in breastfeeding rates across the first year of life. For current rates, we used national- and state-level data²³ for breastfeeding initiation, exclusive breastfeeding at 2 days, 3 months, and 6 months, and any breastfeeding at 6 and 12 months (Appendix Tables A3 and A4); users entered “future” rates at the same time points. These data were used to interpolate differences in breastfeeding rates at each of the 20 time points modeled. For each disease, the difference in breastfeeding rate at each time point was multiplied by marginal effect estimate at that time point, and these values were summed. CIs were calculated by summing CIs for each time point in quadrature.

Finally, the summed estimate of change in disease outcomes was weighted by the population of interest. For the United States, after our prior study,² we used the total population of women who were aged 15 in 2002, 1,994,060, and a total of 3,952,841 births over their lifetime. For the outcome of NEC, we modeled a total of 56,252 very low birth weight (VLBW) births. For the outcome of childhood obesity, we used the number of NHW women in 2002, which was 1,259,336.

To estimate the effect of changes in breastfeeding rates for each of the 50 states in the United States, we used census data to determine the population of 15-year-old women in each state in 2002, and we used 2012 vital statistics data to determine the number of overall births and the number of VLBW births in each state (Appendix Table A5). Detailed data on births by maternal age are not available for each state; therefore, our models use national data to estimate age at each birth and cumulative lifetime parity. For each state, we weighted results by the ratio of the number of births in that state in 2012 to the state population of women who were aged 15 in 2002.

We similarly weighted NEC by the ratio of the number of VLBW births in that state to the number of women who were aged 15 in 2002. Of note, because state-level breastfeeding rates for VLBW infants are not available, we assumed that, under current conditions, VLBW infants were breastfed at the same rate as term infants in the same state. For child obesity, we weighted results by the ratio of state NHW births in 2012 per 15-year-old NHW women in 2002 to the national number of NHW births in 2012 per 15-year-old NHW women in 2002. All ratios used appear in Appendix Table A5.

Costs were calculated by multiplying number of cases and deaths for each disease by the average cost per case over the cohort's lifetime, estimated in our previous paper (Appendix Table A6). Deaths and costs were summed, with CIs summed in quadrature, to estimate total direct and indirect medical costs, total non-medical costs, and societal costs due to premature death. To validate this approach, we compared results by using our marginal model to results previously published by our group from two sets of Monte Carlo simulations, one under suboptimal conditions, using breastfeeding rates from 2012, and one under optimal conditions, defined as 90% of women achieving exclusive breastfeeding for 6 months. Next, we used our marginal model approach to estimate the impact on disease burden of (1) a 5% increase in breastfeeding rates at each time point, for national disease burden; and (2) achieving Healthy People 2020 (HP 2020) goals for national- and state-level disease burden.

To allow others to calculate differences in health outcomes by state and change in breastfeeding rates, we deployed the calculator as a Javascript-based interactive tool, which is publically available at www.usbreastfeeding.org/saving-calc. The Institutional Review Board of the Cambridge Health Alliance exempted the study.

Results

Using the techniques described, we estimated the marginal impact on cases and costs of a 10% change in breastfeeding rates for each of five maternal and nine child diseases. We found marked effects on particular time points, reflecting that the literature on which our model is based uses cut points for breastfeeding duration (Figs. 1 and 2). For example, for otitis media, exclusive breastfeeding for >3 months is associated with an odds ratio of 0.5 (95% CI 0.36–0.70). Thus, a 10% change in exclusive breastfeeding at 3 months is associated with a difference of 5,423.41 cases per 100,000 mothers, whereas a 10% change in exclusive breastfeeding at 4 months is not associated with a meaningful difference.

FIG. 1.

Maternal outcomes, difference, 95% CI for change number of cases per 100,000 mothers for a 10% absolute change in any breastfeeding, months 0–12, and exclusive breastfeeding, months 0–6 shaded. CI, confidence interval.

FIG. 2.

Child outcomes, difference, 95% CI for change number of cases per 100,000 mothers for a 10% absolute change in any breastfeeding, months 0–12, and exclusive breastfeeding, months 0–6 shaded.

When we compared results from our calculated model with our previously published Monte Carlo simulation model (Table 1), we found that our effect estimates were very similar for common conditions, such as otitis media and gastrointestinal illness, although CIs were wider. For less common conditions (e.g., Crohn's/ulcerative colitis and sudden infant death syndrome), differences were more pronounced, and for ovarian cancer, which was not statistically significant in our previously published models, results were markedly different, reflecting the instability of these effect estimates. NEC disease burden was markedly lower in our calculated model, because we used population-level data for baseline breastfeeding rates, rather than VLBW-specific breastfeeding rates.

Table 1.

Comparison of Calculated Marginal Impact Model Versus Monte Carlo Simulation Model

	Calculator		Monte Carlo simulation		Calculator		Monte Carlo simulation
Disease	Cases	95% CI	Cases	95% CI	Deaths	95% CI	Deaths	95% CI
Maternal
Pre-menopausal OC	−22	(−482 to 451)	22	−71 to 112	−29	(−316 to 280)	8	−58 to 71
Breast cancer	4157	(−357 to 8938)	5,023	3,965 to 6,021	790	(−1072 to 2827)	838	434 to 1,245
Hypertension	29986	(21425 to 39743)	35,982	34,122 to 38,144	260	(−689 to 1415)	322	98 to 543
Diabetes	10504	(2784 to 19495)	12,320	10,537 to 14,162	410	(−1025 to 1964)	473	154 to 789
Myocardial infarction	7258	(2420 to 12632)	8,487	7,520 to 9,583	838	(−422 to 2364)	986	677 to 1,295
Total maternal					2268	(−577 to 5482)	2,619	1,978 to 3,259
Child
Acute lymphoblastic leukemia	226	(−388 to 900)	185	49 to 309	41	(−210 to 344)	37	−22 to 91
Crohn's disease	184	−248 to 739	145	31 to 249
Ulcerative colitis	138	(−436 to 759)	136	18 to 263
Sudden infant death syndrome					475	(45 to 934)	492	395 to 588
Otitis media	597150	(573127 to 627092)	601,825	596,885 to 609,362
Gastrointestinal infection	2,539,282	(2,493,851 to 2,593,486)	2,558,629	2,554,934 to 2,577,865
Obesity (NHW only)	39689	(35880 to 43161)	45,298	44,408 to 46,353
Lower respiratory tract infection requiring hospitalization	20563	(17302 to 24883)	20,900	20,014 to 21,836	33	(−108 to 220)	40	4 to 73
Necrotizing enterocolitis	572	(85 to 1124)	1,355	1,237 to 1,489	76	(−114 to 288)	190	144 to 239
Total child					626	(75 to 1244)	721	543 to 899

CI, confidence interval; NHW, non-Hispanic White; OC, ovarian cancer.

We found that a 5% point increase in national breastfeeding rates at all time points (Table 2) was associated with statistically significant differences in child infectious morbidity, including otitis media (101,952 cases, 95% CI 77,929–131,894 cases) and gastrointestinal infection (236,073 cases, 95% CI 190,643–290,278 cases). Associated medical cost differences were $31,784,763 (95% CI $24,295,235–$41,119,548) for otitis media and $12,588,848 ($10,166,203–$15,479,352) for gastrointestinal infection. A 5% absolute increase in breastfeeding was also associated with a reduction in child obesity (4,188 cases, 95% CI 379–7,659). For differences in maternal health outcomes and less common child outcomes, CIs were wide.

Table 2.

Effects of a 5% Absolute Increase in Rates of Any and Exclusive Breastfeeding at Each Time Point on Total Disease Burden in the United States, Based on Marginal Model

	Cases		Deaths		Total medical costs		Indirect medical costs		Death costs
Disease	N	95% CI	N	95% CI	Cost ($)	95% CI ($)	Cost ($)	95% CI ($)	Cost ($)	95% CI ($)
Maternal
Pre-menopausal OC	0	−460 to 473	−3	−290 to 306	37,597	−52,237,080 to 53,727,733	8,610	−11,962,292 to 12,303,652	12,473,416	−1,409,855,704 to 1,487,638,964
Breast cancer	519	−3,995 to 5,300	93	−1,769 to 2,130	12,250,769	−94,310,097 to 125,102,153	2,805,427	−21,597,018 to 28,648,400	263,007,911	−4,985,550,490 to 6,003,997,975
Hypertension	3,877	−4,684 to 13,634	33	−916 to 1,188	31,765,833	−38,372,289 to 111,698,073	4,174,714	−5,042,945 to 14,679,531	85,488,548	−2,398,823,018 to 3,111,888,789
Diabetes	1,297	−6,423 to 10,288	52	−1,383 to 1,606	127,013,144	−629,198,619 to 1,007,841,750	33,554,871	−166,224,361 to 266,255,910	164,767,704	−4,395,810,939 to 5,103,683,176
Myocardial infarction	807	−4,030 to 6,181	89	−1,170 to 1,616	75,815,270	−378,473,167 to 580,449,117	3,474,708	−17,345,894 to 26,602,702	213,121,771	−2,808,369,671 to 3,876,625,547
Total maternal			264	−2,581 to 3,477	246,882,613	−646,001,159 to 1,272,807,357	44,018,330	−158,883,090 to 279,844,656	713,912,518	−7,385,478,484 to 9,779,371,604
Child
Acute lymphoblastic leukemia	27	−586 to 701	5	−246 to 308	3,913,409	−84,329,861 to 100,866,660	283,820	−6,116,033 to 7,315,367	41,424,969	−1,887,500,351 to 2,366,763,173
Crohn's disease	77	−355 to 632			753,005	−3,490,162 to 6,217,857	13,017	−60,333 to 107,485
Ulcerative colitis	68	−505 to 690			500,270	−3,692,755 to 5,045,044	31,482	−232,384 to 317,483
Sudden infant death syndrome			93	−337 to 552					919,093,671	−3,333,339,387 to 5,453,311,627
Otitis media	101,952	77,929 to 131,894			31,784,763	24,295,235 to 41,119,548	17,348,485	13,260,616 to 22,443,517
Gastrointestinal illness	236,073	190,643 to 290,278			12,588,848	10,166,203 to 15,479,352	64,254,436	51,889,071 to 79,007,783
Obesity (NHW only)	4,188	379 to 7,659			12,224,498	1,105,871 to 22,357,344	986,123	89,208 to 1,803,517
Lower respiratory tract infection	1,924	−1,338 to 6,244	3	−138 to 190	10,441,431	−7,259,892 to 33,889,434	1,850,224	−1,286,454 to 6,005,214	29,993,516	−1,362,395,301 to 1,880,789,700
Necrotizing enterocolitis	141	−346 to 694	20	−171 to 232	3,412,057	−8,345,413 to 16,746,287	67,434	−164,934 to 330,964	197,262,686	−1,698,465,604 to 2,300,446,169
Total child			121	−430 to 739	75,618,282	−16,357,920 to 177,482,067	84,835,021	69,957,412 to 102,474,621	1,187,774,842	−4,040,654,385 to 7,002,855,024

When we considered the national impact of achieving HP 2020 goals, we similarly found statistically significant differences in otitis media (49,642 cases, 95% CI 25,619 – 79,584) gastrointestinal infection (148,362 cases, 95% CI 102,931, 202,566), and obesity (6,266 cases, 95% CI 2,457 – 9,737). For state-level analyses, the impact of attaining HP 2020 goals varied, depending on population size, birth rate, and current breastfeeding rates. For Mississippi, which had the lowest breastfeeding initiation rate in the United States in 2012, achieving HP 2020 goals was associated with a statistically significant difference in maternal hypertension (236 cases, 95% CI 146–339) and diabetes (83 cases, 95% CI 1–177), and with total maternal direct and indirect medical costs of $15,216,972 (95% CI $5,812,423–$26,022,813).

For child outcomes in Mississippi, achieving HP 2020 goals was associated with differences in otitis media (4,742 cases, 95% CI 4,506–5,035), gastrointestinal infection (9,550 cases, 95% CI 9,105–10,081 cases), and hospitalization for lower respiratory tract infection (78 cases, 95% CI 46–120 cases). Achieving HP 2020 goals would also avert 10 cases of NEC (95% CI 3–18 cases) and 224 cases of child obesity (95% CI 188–257). Child medical costs averted totaled $3,607,151 (95% CI $2,698,500–$4,614,066). Finally, achieving HP 2020 goals in Mississippi would avert seven child deaths (95% CI 1–14) with societal savings of $69,461,935 (95% CI $14,646,198–130,408,942).

By contrast, Washington had the highest rate of breastfeeding initiation in 2012, at 93.6%, and it exceeds HP 2020 goals on every metric. As a result, if rates were to fall to HP 2020 target levels in Washington, disease burden would increase, with 4,132 excess otitis media cases (95% CI 3,471–4,662) and 4,800 excess gastrointestinal infections (95% CI 3,603–5,803). Point estimates for all other conditions indicated a higher disease burden with HP 2020 breastfeeding rates, but CIs were wide.

Discussion

We used a Monte Carlo simulation to estimate the change in disease rates associated with changes in rates of any and exclusive breastfeeding; we used these results to create a calculator to estimate the effects of changes in breastfeeding on population health. Our results suggest that even modest increases in breastfeeding rates are associated with changes in disease burden and costs, particularly for childhood otitis media and gastrointestinal infections, which are common and for which breastfeeding is associated with a marked reduction in risk.

Evidence suggests that the 5% difference in breastfeeding rates modeled can be achieved with individual-level interventions to support healthy breastfeeding mothers with healthy babies.²⁴ In a recent Cochrane review of 52 studies among 56,451 mother-infant pairs from 21 countries, Renfrew et al.⁷ found that breastfeeding support increased rates of any breastfeeding at 6 months from 45% to 49%, and exclusive breastfeeding at 6 months from 17% to 26%. Patnode et al.²⁵ similarly systematically reviewed interventions to support breastfeeding, finding that interventions increased rates of exclusive breastfeeding at 6 months, Relative Risk (RR) 1.16 (95% CI, 1.02–1.32), with absolute differences ranging from a decrease of 2.9% to an increase of 21%.

Further, there is evidence that systems-level factors, such as maternity care practices,²⁶ formula marketing,²⁷ and the composition of the Women, Infants and Children food package,²⁸ affect the likelihood that mother-infant dyads will be able to breastfeed exclusively. We estimated that a 5% absolute increase in breastfeeding rates would save $31,784,763 in otitis media medical costs (95% CI $24,295,235–$41,119,548) and $12,588,848 in gastrointestinal infection costs (95% CI $10,166,203–$15,479,352), suggesting that investing in existing strategies is cost effective.

Our modeling approach has several important limitations. To generate marginal effect estimates, we performed 5,000 replications of 100,000 women per replication, half the number of replications in our prior work. The higher the number of replications, the greater the precision of estimates, and the narrower the CI around each marginal effect. For rare outcomes, in particular, running 5,000, rather than 10,000, replications resulted in larger CIs. Given the smaller size of state populations, these wider CIs for rare outcomes are more conservative and likely appropriate.

Further, the Monte Carlo simulation models a cohort of women who were aged 15 in 2002 across their lifetimes, assuming a steady-state population. In some regions, it is common for young adults to relocate away from their state of birth after age 15; in other regions, large numbers of reproductive-aged women move into a state after age 15. These population shifts likely drive differences in the ratio of births per year to the number of women aged 15 in 2002. Nationally, the ratio is 1.88; whereas in the District of Columbia, the ratio is 3.60, and in New Hampshire, it is 1.35. We used these ratios to weight the number of child cases in our simulation results; however, our estimates for maternal disease burden did not account for migration of women into or out of a state.

We also used national birth rate data to model the timing of births for cohort participants. Birth timing affects costs, because child health costs are discounted, and the later in life a birth occurs, the greater the discounting of costs for that child. In states where women tend to give birth at younger ages than the national average, our model, therefore, overestimates cost; whereas in states where women tend to give birth at older ages than the national average, our model underestimates cost.

Our estimates of cases and costs for NEC were limited by lack of state-level data on breastfeeding rates among VLBW infants. In our prior publications,^2,15 we used data from Rush Hospital to estimate breastfeeding rates for VLBW infants. At 36 weeks post-menstrual age, 24% of infants at Rush receive exclusive human milk. National rates of exclusive breastfeeding at 3 months are 43.3%, with state rates ranging from 23% to 56%. In the absence of any information on the extent to which state breastfeeding rates are correlated with breastfeeding rates for VLBW infants, we did not model different breastfeeding rates for VLBW infants in our calculator. To the extent that overall breastfeeding rates are higher than VLBW rates, modeled differences between current and future rates would be smaller, and our calculator model likely underestimates reductions in NEC that are attributable to achieving a given breastfeeding rate.

Conclusion

We created a web-based calculator to estimate the expected impact of changes in breastfeeding rates on population health outcomes for each of the 50 U.S. states. This tool allows policy makers and healthcare payers to estimate the impact of changes in breastfeeding rates on maternal and child health and healthcare costs, subject to the limitations in our modeling approach described earlier. Our results suggest that even a 5% increase in breastfeeding rates could substantially impact disease burden for otitis media and gastrointestinal infection in the first year of life, with savings of more than $44,000,000 in direct medical costs for these two conditions alone. This tool can inform policymakers about the expected return on investments to enable more women to achieve their infant feeding goals.

Footnotes

Acknowledgment

This research was funded by the W.K. Kellogg Foundation. The funder had no role in the study design or article preparation.

Disclosure Statement

No competing financial interests exist.

Appendix

Appendix Table A6.

Costs per Case

Disease	2% Discount direct medical ($)	2% Discount indirect medical ($)	2% Discount total medical ($)	2% Discount indirect non-medical ($)	2% Discount death ($)
Pre-menopausal ovarian cancer	96,543.61	17,037.10	113,580.72	26,009.99	4,864,140.94
Breast cancer	20,065.27	3,540.93	23,606.19	5,405.82	2,818,223.32
Hypertension	6,963.97	1,228.92	8,192.90	1,076.72	2,619,232.48
Diabetes	83,265.99	14,694.00	97,959.99	25,879.49	3,178,292.58
Myocardial infarction	79,823.00	14,086.44	93,909.44	4,303.99	2,399,441.00
Acute lymphoblastic leukemia	125,040.33	18,756.05	143,796.38	10,428.85	7,677,080.96
Crohn's disease	8,364.97	1,476.17	9,841.14	170.12
Ulcerative colitis	6,214.07	1,096.60	7,310.67	460.06
Sudden infant death syndrome					9,886,863.68
Otitis media	265.00	46.76	311.76	170.16
Gastrointestinal illness	48.92	4.40	53.33	272.18
Obesity (NHW only)	2,481.16	437.85	2,919.02	235.47
Lower respiratory tract infection requiring hospitalization	4,613.64	814.17	5,427.81	961.81	9,895,677.23
Necrotizing enterocolitis	20,883.35	3,257.94	24,141.29	477.11	9,936,233.24

References

Wolf

. Low breastfeeding rates and public health in the United States. Am J Public Health, 2003; 93:2000–2010.

Bartick

, Schwarz

, Green

, et al. Suboptimal breastfeeding in the United States: Maternal and pediatric health outcomes and costs. Matern Child Nutr, 2017; 13:1–13.

Stuebe

, Horton

, Chetwynd

, et al. Prevalence and risk factors for early, undesired weaning attributed to lactation dysfunction. J Womens Health (Larchmt), 2014; 23:404–412.

American Academy of Pediatrics. Breastfeeding and the use of human milk. Pediatrics, 2012; 129:e827–e841.

American College of Obstetricians and Gynecologists. Optimizing support for breastfeeding as part of obstetric practice. Commitee Opinion No. 658. Obstet Gynecol, 2016; 127:e86–e92.

Montgomery

, Dermer

, Eglash

, et al. Family physicians supporting breastfeeding. Updated. 2016. Available at www.aafp.org/about/policies/all/breastfeeding-support.html (accessed October 7, 2016).

Renfrew

, McCormick

, Wade

, et al. Support for healthy breastfeeding mothers with healthy term babies. Cochrane Database Syst Rev, 2012; 5:CD001141.

Patnode

, Henninger

, Senger

, et al. Primary Care Interventions to Support Breastfeeding: An Updated Systematic Review for the U.S. Preventive Services Task Force. AHRQ Publication No. 15-05218-EF-1. Rockville, MD: Agency for Healthcare Research and Quality, 2016.

Amitay

, Keinan-Boker

. Breastfeeding and childhood leukemia incidence: A meta-analysis and systematic review. JAMA Pediatr, 2015; 169:e151025.

10.

, Chung

, Raman

, et al. Breastfeeding and Maternal and Infant Health Outcomes in Developed Countries. Evidence Report/Technology Assessment Number 153. Vol AHRQ Publication No. 07-E007. Rockville, MD: Agency for Healthcare Research and Quality, 2007.

11.

Hornell

, Lagstrom

, Lande

, et al. Breastfeeding, introduction of other foods and effects on health: A systematic literature review for the 5th Nordic Nutrition Recommendations. Food Nutr Res, 2013; 57:20823.

12.

Klement

, Cohen

, Boxman

, et al. Breastfeeding and risk of inflammatory bowel disease: A systematic review with meta-analysis. Am J Clin Nutr, 2004; 80:1342–1352.

13.

Wolters

, Russel

, Sijbrandij

, et al. Crohn's disease: Increased mortality 10 years after diagnosis in a Europe-wide population based cohort. Gut, 2006; 55:510–518.

14.

Duijts

, Jaddoe

, Hofman

, et al. Prolonged and exclusive breastfeeding reduces the risk of infectious diseases in infancy. Pediatrics, 2010; 126:e18–e25.

15.

Colaizy

, Bartick

, Jegier

, et al. Impact of optimized breastfeeding on the costs of necrotizing entercolitis in extremely low birthweight infants. J Pediatr, 2016; 175:100–105.

16.

Hauck

, Thompson

, Tanabe

, et al. Breastfeeding and reduced risk of sudden infant death syndrome: A meta-analysis. Pediatrics, 2011; 128:103–110.

17.

Grummer-Strawn

, Mei

. Does breastfeeding protect against pediatric overweight? Analysis of longitudinal data from the Centers for Disease Control and Prevention Pediatric Nutrition Surveillance System. Pediatrics, 2004; 113:e81–e86.

18.

Danforth

, Tworoger

, Hecht

, et al. Breastfeeding and risk of ovarian cancer in two prospective cohorts. Cancer Causes Control, 2007; 18:517–523.

19.

Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer and breastfeeding: Collaborative reanalysis of individual data from 47 epidemiological studies in 30 countries, including 50302 women with breast cancer and 96973 women without the disease. Lancet, 2002; 360:187–195.

20.

Stuebe

, Rich-Edwards

, Willett

, et al. Duration of lactation and incidence of type 2 diabetes. JAMA, 2005; 294:2601–2610.

21.

Stuebe

, Schwarz

, Grewen

, et al. Duration of lactation and incidence of maternal hypertension: A longitudinal cohort study. Am J Epidemiol, 2011; 174:1147–1158.

22.

Stuebe

, Michels

, Willett

, et al. Duration of lactation and incidence of myocardial infarction in middle to late adulthood. Am J Obstet Gynecol, 2009; 200:138.e1–e8.

23.

Centers for Disease Control and Prevention. Breastfeeding among U.S. children born 2000–2013, CDC National Immunization Survey. Available at www.cdc.gov/breastfeeding/data/nis_data/index.htm (accessed September 8, 2016).

24.

Bonuck

, Stuebe

, Barnett

, et al. Effect of primary care intervention on breastfeeding duration and intensity. Am J Public Health, 2014; 104 Suppl 1:S119–S127.

25.

Patnode

, Henninger

, Senger

, et al. Primary care interventions to support breastfeeding: Updated evidence report and systematic review for the US Preventive Services Task Force. JAMA, 2016; 316:1694–1705.

26.

Declercq

, Labbok

, Sakala

, et al. Hospital practices and women's likelihood of fulfilling their intention to exclusively breastfeed. Am J Public Health, 2009; 99:929–935.

27.

Rosenberg

, Eastham

, Kasehagen

, et al. Marketing infant formula through hospitals: The impact of commercial hospital discharge packs on breastfeeding. Am J Public Health, 2008; 98:290–295.

28.

Langellier

, Chaparro

, Wang

, et al. The new food package and breastfeeding outcomes among women, infants, and children participants in Los Angeles County. Am J Public Health, 2014; 104 Suppl 1:S112–S118.