Background: Heat-treated expressed breast milk is recommended by the World Health Organization as an option to reduce vertical HIV transmission in resource-poor regions. Flash-heat (FH) is a low technology pasteurization method developed for home use, but its effect on quantity and quality of breast milk immunoglobulins is unknown.
Objective: To evaluate FH's effect on breast milk immunoglobulin levels and antigen-binding capacity.
Design/Methods: Fifty HIV+ mothers in South Africa provided breast milk. Part of each sample served as an unheated control; the remainder was flash-heated. Total and antigen-specific immunoglobulin A (IgA) and immunoglobulin G (IgG) were measured by enzyme-linked immunosorbent assay. Paired t test was performed on log-transformed data.
Results: FH significantly decreased total IgA and IgG concentrations [geometric mean (geometric SD) 318.0 (1.9) vs. 398.2 (1.9) μg/mL and 89.1 (2.7) vs. 133.3 (2.5) μg/mL, P < 0.001 each]. Similar decreases in anti–HIV-1 gp120 IgG, anti–pneumococcal polysaccharide, and anti–poliovirus IgA occurred (P < 0.001 each). Although the latter was most affected, FH retained 66% of the antigen-binding ability. In contrast, binding capacity of IgA and IgG to influenza increased after FH (P = 0.029 and 0.025, respectively).
Conclusions: Most breast milk immunoglobulin activity survives FH, suggesting flash-heated breast milk is immunologically superior to breast milk substitutes. Clinical significance of this decreased immunoglobulin activity needs evaluation in prospective trials.
From the *Department of Pediatrics, University of California Davis Medical Center, Sacramento, CA; †Program in International and Community Nutrition, Department of Nutrition, University of California Davis, Davis, CA; ‡Department of Microbiology; University of Alabama at Birmingham, Birmingham, AL; §Department of Paediatrics and Child Health, Nelson R. Mandela School of Medicine, University of Kwazulu-Natal, Durban, South Africa; ∥School of Dietetics and Human Nutrition, McGill University, Montreal, Quebec, Canada; and ¶Division of Epidemiology, School of Public Health, University of California Berkeley, Berkeley, CA.
Received for publication July 30, 2008; accepted January 26, 2009.
Supported by funding by the National Institute of Child Health and Human Development (Grant #HD051473-01), the University of California, Davis Children's Miracle Network, the Thrasher Research Fund, and the James B. Pendleton Charitable Trust.
Correspondence to: Caroline J. Chantry, MD, Department of Pediatrics, University of California Davis Medical Center, 2516 Stockton Blvd. Sacramento, CA 95817 (e-mail: firstname.lastname@example.org).
Prolonged breastfeeding accounts for up to 40% of maternal to child transmission of HIV in resource-poor regions of the world.1 Multiple studies, however, document that HIV-free infant survival is not improved in many of these areas by use of breast milk substitutes.2–5 When infants are not breastfed in these regions, an increase in malnutrition6,7 and morbidity and mortality from diarrhea8–10 result. Accordingly, ways to decrease maternal to child transmission during breastfeeding could potentially improve HIV-free child survival.
The World Health Organization recommends pasteurization of breast milk as a modification to breastfeeding in this setting.11,12 We have previously described a low “tech” method of pasteurization, flash-heat, which mothers can use in their homes, and documented that this method can successfully inactivate cell-free HIV in naturally infected human milk13 and in high-titer ‘spiked’ breast milk.14 Before subjecting this novel pasteurization method to clinical trial, it was necessary to ascertain the effect on breast milk immunoglobulins to ensure that the milk would continue to offer passive immunoprotection.
Flash-heat was designed to mimic commercial flash pasteurization, a high-temperature short-time (HTST) pasteurization method. As a general principal, HTST methods more effectively kill microorganisms while better preserving nutritional food value when compared with low-temperature long-time pasteurization methods.15–17
Effects of low-temperature long-time pasteurization methods on immunoglobulin A (IgA) and immunoglobulin G (IgG) in milk have been extensively studied,18–20 but minimal work has examined effects of HTST methods on breast milk immunoglobulins.21 Furthermore, flash-heat raises and lowers the milk's temperature more slowly than does its high “tech” counterpart, which rapidly heats liquid to 72°C for 15 seconds, and therefore can potentially cause greater harm. The objective of this study was to evaluate the effects of flash-heat treatment on concentrations of breast milk IgA and IgG and on their binding capacity to selected relevant microbial antigens.
Fifty breast milk samples were collected from HIV-infected women in Durban, South Africa, between October and December, 2004. Clinical and demographic characteristics of these women and breast milk collection procedures have been previously described.22 Briefly, mean [SD (range)] maternal age was 25.9 [4.9 (19–40)] years, body mass index was 27.5 [4.3 (20.0–37.5)] kg/m2, and CD4+ cell count was 527 [255 (27–1173)]; mean infant age was 15 [11 (6–68)] weeks.
After aliquotting an unheated control, the remainder of the fresh milk was flash-heated in the laboratory under conditions designed to mimic those in the field. Briefly, 50 mL of milk was placed in an uncovered 16 oz (455 mL) glass food jar, which was then placed in 450 mL of water in a 1:1 Hart brand 1 quart aluminum pan. The water and milk were heated together over a butane stove burner, used to imitate the intense heat of a fire, until the water reached 100°C and was at a rolling boil. The jar of breast milk was then immediately removed from the water bath and allowed to cool to 37.0°C. Time–temperature curve of the milk is shown in Figure 1. The breast milk typically reached a peak temperature of 72.9°C and was above 56.0°C for 6 minutes 15 seconds. Samples were stored at −70°C until analysis.
Total and antigen-specific IgA and IgG levels were measured in treated and untreated milk samples by enzyme-linked immunosorbent assay (ELISA). High binding capacity polystyrene 96-microwell ELISA plates (Nalge Nunc, Rochester, NY) were coated overnight either with 1 μg/mL of F(ab')2 fragment of goat IgG specific for human IgA or IgG isotypes (Jackson ImmunoResearch Laboratories, West Grove, PA) or with the following microbial antigens: 1 μg/mL rgp120 of HIV-1 (prepared and purified at University of Alabama, Birmingham, AL); inactivated trivalent influenza viruses purified subvirion antigens at 0.3 μg/mL hemagglutinin of each influenza virus types A (H3N2 and H1N1) and B (Flushield vaccine; Wyeth, Marietta, PA); inactivated poliovirus (University of Alabama, Birmingham, AL) at 2 μg/mL; 23-valent pneumococcal polysaccharide vaccine (Merck, West Point, PA) diluted 1:100; and Salmonella typhosa lipopolysaccharide (Sigma, St. Louis, MO) at 5 μg/mL. Plates were blocked with 5% goat serum in phosphate-buffered saline containing 0.05% Tween 20 for 2 hours at room temperature. After washing the plates, serial 2-fold dilutions of samples, standards [a pool of human sera calibrated for immunoglobulin isotype levels (The Binding Site, Birmingham, UK)], and positive controls (human sera with known positivity for each of the antigens analyzed) were incubated overnight at 4°C. The captured antibodies were then detected by sequential addition of: (a) biotin-conjugated F(ab')2 fragment of goat IgG specific for human IgA or IgG antibodies (BioSource, Camarillo, CA); (b) horseradish peroxidase–labeled ExtrAvidin (Sigma); (c) the chromogenic substrate for horseradish peroxidase: orthophenylenediamine and 0.0075% hydrogen peroxide (Sigma). The color reaction was stopped with 1 M sulfuric acid, and absorbance at 490 nm was read in a Bio-Kinetics reader (Bio-Tek Instruments, Winooski, VT).
Results were calculated for total and antigen-specific IgA and IgG by interpolating the optical densities of samples on calibration curves constructed from standardized sera using Delta Soft computer program (BioMettalics, Princeton, NJ). Data were log transformed to achieve normal distribution and analyzed using paired t test.
In the 50 samples analyzed, flash-heat induced a statistically significant decrease in total IgA [geometric mean (SD) 318.0 (1.9) vs. 398.2 (1.9) μg/mL, P < 0.001] and IgG [89.1 (2.7) vs. 133.3 (2.5) μg/mL, P < 0.001] concentrations, which corresponded to a 20% (95% CI 15 to 25) and 33% (27 to 39) reduction, respectively. Similar decreases were observed in levels of HIV-1 gp120-specific IgG [26% (18 to 33), P < 0.001] and in anti–pneumococcal polysaccharide and anti–poliovirus IgA [30% (21 to 38) and 34% (26 to 41), P < 0.001 each]; the anti–poliovirus IgA being the most affected of all the antibodies measured. In contrast, the amount of IgA and IgG binding to influenza viruses increased after flash-heat by 13% (2 to 26), P = 0.029 and 15% (2 to 31), P = 0.025, respectively (Table 1). The increase in anti–Salmonella lipopolysaccharide IgA of 9% (−2 to 21) did not reach statistical significance, P = 0.13.
The majority of IgA and IgG activity in these samples survived the flash-heat treatment and, in the case of anti–influenza immunoglobulins, the antigen-binding capacity of both increased. These results suggest that infants receiving flash-heated breast milk would still receive substantial passive immunoprotection. The IgA concentration measured after flash-heating was nearly identical to that reported by Goldblum et al21 who observed a decrease in IgA concentration from 0.37 to 0.30 mg/mL (19%) after heating breast milk to 72°C for 15 seconds; this difference was not statistically significant in that study, possibly due to the fewer samples analyzed. These results suggest that our “low-tech” version of HTST requiring longer time to reach 72°C did not substantively increase the immunoglobulin denaturation compared with the “high-tech” version. A similar proportion of IgA survives the traditional Holder pasteurization used in breast milk banks (62.5°C for 30 minutes). Gibbs et al20 detected a 21% loss of IgA with Holder pasteurization of drip milk and Ford et al19 noted 20% reduction in IgA titer with this method. However, very little IgA has been noted to survive higher temperatures, with greater than 80% reduction at 87°C for 1 second21 and essentially total destruction with boiling.23 IgG is apparently more heat sensitive than IgA, as noted by 33% vs. 20% destruction by the flash-heat method in our study; greater destruction of IgG relative to IgA was also noted by Evans et al24 when heating to 73°C for 30 minutes.
Most importantly, the antibody activity of IgA and IgG, measured as antigen-binding capacity, was also predominantly preserved with the flash-heat method, with postheat activity ranging from 66% of that measured preheat for anti–poliovirus IgA to 115% for anti–influenza virus IgA. It is unclear why the influenza virus-specific antibodies slightly increased in our samples after pasteurization and if this would correlate clinically to enhanced protection. Chen and Allen25 reported that 61.1% of Escherichia coli–specific IgA activity was retained after HTST pasteurization of human milk, a percentage slightly lower than that of polio-specific IgA in our study, the most diminished by flash-heat. Again, this suggests that there is no greater impact on antibody activity when the milk is heated more slowly during the flash-heat method vs. “high-tech” methodologies. It is reassuring to note that Carbonare et al26 in their study of Holder pasteurization found that although concentration and anti-enteropathogenic E. coli activity of IgA in colostrum were reduced, the remaining IgA was sufficient to effectively inhibit bacterial adhesion to HEp-2 cells.
IgA concentration in breast milk is known to decline as the infant ages. The geometric mean concentration of IgA in our unheated milk samples from the HIV-infected women was 0.40 mg/mL (arithmetic mean was 0.46 mg/mL) at a mean postpartum time of 15 weeks. Previously reported means are somewhat variable: 0.71 mg/mL at 10 weeks postpartum,27 0.50 mg/mL at 12 weeks,28 and 0.57 mg/mL at a mean of 24 weeks.29 As malnourished women have less IgA in their colostrum,30 it is not surprising that the samples from our cohort had slightly lower levels of IgA than the samples from the above reports.27–29 Although no women in our study were underweight, they were from an impoverished region and may have had micronutrient deficiencies. HIV infection may also contribute to alterations in IgA content. To the best of our knowledge, effects of HIV infection per se on breast milk composition have not yet been studied. Alternatively, the differences may be due to laboratory methodologies as the original reports were all performed more than 25 years ago and used radial immunodiffusion rather than ELISA to measure immunoglobulin levels. In contrast to the lower concentrations of IgA, IgG levels in the milk from HIV-infected mothers in this study were substantially higher than those previously reported in mature milk (30 mg/mL),31 presumably because of the hypergammaglobulinemia frequently found in HIV-infected individuals.32
The current study was limited to laboratory measurement of IgA and IgG concentrations and binding capacities for 5 representative antigens. Antibody activity for other antigens may have been more or less affected. Also, this study reports the impact of flash-heat on immunoglobulins only, whereas many other breast milk components provide anti-infective activity and immunoprotection. Work is currently underway to evaluate flash-heat's effect on bioactivity of key breast milk proteins.
We anticipate that because flash-heated milk retains most antibody specificity for the microbial antigens tested, it will confer similar protection from infection for the infant as would unheated milk. Correlation with clinical data from field trials will be necessary. In conclusion, this study documents that the majority of breast milk IgA and IgG survive flash-heat treatment and retain the ability to bind specific antigens, suggesting that this method would be immunologically superior to boiling milk or using breast milk substitutes. The clinical significance of the observed decrease in antibody activity to some antigens will need further evaluation in prospective clinical trials. Flash-heat may be a safe and affordable method for home pasteurization for HIV+ mothers in developing countries, of particular value during times of greater risk for HIV transmission, such as during episodes of infant oral thrush or maternal mastitis or upon addition of complementary foods.
The authors would like to acknowledge the mothers who participated in this study and the Cato Manor Clinic staff for their time and dedication.
1. Nduati R, John G, Mbori-Ngacha D, et al. Effect of breastfeeding and formula feeding on transmission of HIV-1: a randomized clinical trial. JAMA
2. Coutsoudis A, Pillay K, Kuhn L, et al. Method of feeding and transmission of HIV-1 from mothers to children by 15 months of age: prospective cohort study from Durban, South Africa. AIDS
3. Kuhn L, Aldrovandi GM, Sinkala M, et al. Effects of early, abrupt weaning for HIV-free survival of children in Zambia. N Engl J Med
4. Mbori-Ngacha D, Nduati R, John G, et al. Morbidity and mortality in breastfed and formula-fed infants of HIV-1- infected women: a randomized clinical trial. JAMA
5. Becquet R, Bequet L, Ekouevi DK, et al. Two-year morbidity-mortality and alternatives to prolonged breast-feeding among children born to HIV-infected mothers in Cote d'Ivoire. PLoS Med
6. Becquet R, Leroy V, Ekouevi DK, et al. Complementary feeding adequacy in relation to nutritional status among early weaned breastfed children who are born to HIV-infected mothers: ANRS 1201/1202 Ditrame Plus, Abidjan, Cote d'Ivoire. Pediatrics
7. Johnson W, Alons C, Fidalgo L, et al. The challenge of providing adequate infant nutrition following early breastfeeding cessation by HIV-positive, food-insecure Mozambican mothers. Presented at: XVI International AIDS Conference; August 13–18, 2006; Toronto, Canada.
8. Creek T, Arvelo W, Kim A, et al. Role of infant feeding and HIV in a severe outbreak of diarrhea and malnutrition among young children, Botswana, 2006. Presented at: XIV Conference on Retroviruses and Opportunistic Infections; February 25–28, 2007; Los Angeles, CA.
9. Creek T, Arvelo W, Kim A, et al. A large outbreak of diarrhea among non-breastfed children in Botswana 2006—implications for HIV prevention strategies and child health. Presented at: XIV Conference on Retroviruses and Opportunistic Infections; February 25–28, 2007; Los Angeles, CA.
10. Kourtis AP, Fitzgerald G, Hyde L, et al. Diarrhea in uninfected infants of HIV-infected mothers who stop breastfeeding at 6 months: the BAN study experience. Presented at: XIV Conference on Retroviruses and Opportunistic Infections; February 25–28, 2007; Los Angeles.
11. WHO, UNICEF, United Nations Population Fund, UNAIDS. HIV and infant feeding: a guide for health-care managers and supervisors. 2004; Geneva, Switzerland. Available at: http://whqlibdoc.who.int/hq/2003/9241591234.pdf
. Accessed April 30, 2009.
13. Israel-Ballard K, Donovan R, Chantry C, et al. Flash-heat inactivation of HIV-1 in human milk: a potential method to reduce postnatal transmission in developing countries. J Acquir Immune Defic Syndr
14. Israel-Ballard K, Chantry C, Dewey K, et al. Viral, nutritional, and bacterial safety of flash-heated and pretoria-pasteurized breast milk to prevent mother-to-child transmission of HIV in resource-poor countries: a Pilot Study. J Acquir Immune Defic Syndr
15. Dhar J, Fichtali J, Skura BJ. Pasteurization efficiency of a HTST system for human milk. J Food Sci
16. Morgan JN, Lin FJ, Eitenmiller RR, et al. Thermal destruction of Escherichia coli
and Klebsiella pneumoniae
in human milk. J Food Prot
17. Terpstra FG, Rechtman DJ, Lee ML, et al. Antimicrobial and antiviral effect of high-temperature short-time (HTST) pasteurization applied to human milk. Breastfeed Med
18. Braga L, Palharas D. Effect of evaporation and pasteurization in the biochemical and immunological composition of human milk. J Pediatr
19. Ford JE, Law BA, Marshall VM, et al. Influence of the heat treatment of human milk on some of its protective constituents. J Pediatr
20. Gibbs JH, Fisher C, Bhattacharya S, et al. Drip breast milk: it's composition, collection and pasteurization. Early Hum Dev
21. Goldblum RM, Dill CW, Albrecht TB, et al. Rapid high-temperature treatment of human milk. J Pediatr
22. Israel-Ballard K, Abrams B, Coutsoudis A, et al. Vitamin content of breastmilk from HIV-1 infected mothers before and after flash-heat treatment. J Acquir Immune Defic Syndr
23. Welsh JK, May JT. Anti-infective properties of breast milk. J Pediatr
24. Evans TJ, Ryley HC, Neale LM, et al. Effect of storage and heat on antimicrobial proteins in human milk. Arch Dis Child
25. Chen HY, Allen JC. Human milk antibacterial factors: the effect of temperature on defense systems. Adv Exp Med Biol
26. Carbonare SB, Palmeira P, Silva ML, et al. Effect of microwave radiation, pasteurization and lyophilization on the ability of human milk to inhibit Escherichia coli adherence to HEp-2 cells. J Diarrhoeal Dis Res
27. Mickleson KN, Moriarty KM. Immunoglobulin levels in human colostrum and milk. J Pediatr Gastroenterol Nutr
28. Goldman AS, Garza C, Nichols BL, et al. Immunologic factors in human milk during the first year of lactation. J Pediatr
29. Liebhaber M, Lewiston NJ, Asquith MT, et al. Alterations of lymphocytes and of antibody content of human milk after processing. J Pediatr
30. Miranda R, Saravia NG, Ackerman R, et al. Effect of maternal nutritional status on immunological substances in human colostrum and milk. Am J Clin Nutr
31. Weaver LT, Arthur HM, Bunn JE, et al. Human milk IgA concentrations during the first year of lactation. Arch Dis Child
32. Nagase H, Agematsu K, Kitano K, et al. Mechanism of hypergammaglobulinemia by HIV infection: circulating memory B-cell reduction with plasmacytosis. Clin Immunol