Following the control of major communicable diseases and nutritional problems, allergic disease has emerged as a common cause of illness and even mortality in children. The prevalence of allergic disorders has been reported to have increased in the past 10 to 15 years (1,2). The contributing factors for this increase are unclear. A second reason for the current interest in allergic disease stems from health care costs of treating those with allergy. The treatment of allergic disease, such as eczema, asthma, hay fever, and gastroenteropathy, has considerable economic as well as other consequences that are more difficult or impossible to weigh in monetary terms, for example, absence from school, social isolation, emotional distress, curtailed participation in sports, and others. In Newfoundland, with an annual birthrate of ≈8,400, it costs the health care system ≈$740,800 per year to look after children aged <5 years who had parental history of allergy and had developed allergic disease (3,4)
In high-risk infants with a family history of atopy, many interventions have been attempted to minimize allergic disease. These interventions include exclusive breast-feeding; maternal avoidance of common allergenic foods; use of a hypoallergenic formula; delayed introduction of allergenic solid foods; lessening exposure to dust mites, animals, and tobacco smoke; and avoidance of day care with its attendant risk of infection. Although such interventions have been successful (5,6), it is almost impossible to distinguish and analyze the importance of each individual prophylactic measure. For this reason, it is important to conduct studies in which only one variable, such as the type of infant feeding formula, differs between groups. This recommendation was emphasized by the American Academy of Pediatrics Subcommittee on Nutrition and Allergic Disease (7,8).
We have examined the effects of feeding different infant formulas on the incidence and severity of allergic disease in a prospective randomized double-blind study of 216 high-risk infants with family history of atopy. In previous communications, we reported that the use of a partial whey hydrolysate formula was associated with a lowered incidence of clinical manifestations of allergy until age 18 months, compared with findings in infants fed conventional cow's milk and soy formulas (9,10). We now report follow-up data until 5 years. Besides clinical observations, results of double-blind placebo-controlled food challenges are also reported.
SUBJECTS AND METHODS
Parents were contacted in antenatal clinics or on admission to the Grace General and the St. Clare's Mercy maternity hospitals. Mothers with history of atopic eczema, asthma, or allergic rhinitis or those whose husbands or previous child(ren) had a similar history were recruited and were told the objectives of the study. No attempt was made to grade the severity of atopic disease in the family members. The study was approved by the Human Investigation Committees of the Memorial University of New-foundland and affiliated hospitals. Participating mothers were asked whether they intended to breast-feed exclusively. Mothers who breast-fed their infants were not asked to take any special precautions regarding their own diet. Seventy-two infants with family history of atopy who were breast-fed exclusively for ≥4 months (average 5.1 months) were included in the study. The choice of the 4-month period for exclusive breast-feeding was arbitrary. It has been suggested that the protective effect of breast-feeding increases with its duration. However, if we were to include only those mothers who exclusively breast-fed for 6 months, the recruitment would have been inordinately slow, posing other problems regarding the comparability of this group with the formula groups.
The 216 mothers who intended not to breast-feed were provided one of three coded formulas for their infants: cow's milk, partial whey hydrolysate (NAN/HA, available in the U.S. and Canada under the trademark Good Start), conventional cow's milk (Similac), or soy-based formula (Isomil). There were 72 infants in each formula group. Mothers were not aware of the kind of formula given. Formula-fed infants were fed exclusively on one assigned formula for the entire first 6 months of life. Infants were fed on demand and ad libitum.
The study groups did not differ from one another in terms of several baseline characteristics (Table 1)(9), including family history of atopy, the proportion of infants with cord blood IgE levels ≥0.7 U/ml, and geometric mean cord blood IgE concentration. In addition, the groups did not differ in terms of various confounding factors, such as the presence of household pets, parental smoking, birth weight, parity, gender, socioeconomic status, or maternal education. Day-care facilities were used equally in the three formula groups. None of the breast-fed infants was in day care until after age 6 months.
After age 6 months, the infants were fed according to the general advice given to all families with history of allergy, including stepwise introduction of one new food each week, beginning with vegetables, rice cereal, and fruits. It was advised that egg and fish not be given until ≥18 months, and peanut products not until age 3 years.
Infants were examined by one of several physicians, who were not aware of the kind of formula being given to the infants. Although the formula code was broken for analysis of data at earlier ages, it was known only to the principal investigator. The clinical assistants who saw the study subjects did not know the identity of the formulas. If physical examination found chronic or relapsing dermatitis with skin areas of scaly, erythematous, and itchy rash primarily affecting the face, scalp, area behind the ears, extensor areas, and flexural folds, atopic eczema was diagnosed based on major and minor features (11,12). An eczema score was calculated. The score is based on the extent (body skin surface divided into 20 parts), type (erythema, scaling, lichenification), and severity (each item scored from 0 [absent] to 3 [severe]) of skin involvement. The maximum possible score is 180.
Asthma was defined (13-15) as recurrent episodes of wheezing and/or cough lasting >24 h, diffuse rhonchi and/or rales heard on auscultation of the chest, symptoms reversed with bronchodilators, and when there was no infectious cause likely based on clinical features and appropriate laboratory investigations including complete blood count, radiograph of the chest, and nasopharyngeal swab for respiratory syncytial virus antigen. A minimum of three such episodes in a 12-month period was considered to suggest a diagnosis of asthma.
Skin prick tests were conducted by a standard technique using commercial antigens (milk, soy, egg, peanut, fish, others) in a concentration of 20% wt/vol (Bencard Inc., Ontario, Canada). Double-blind placebo-controlled food challenges (DBPCFC) were performed according to standard protocol (Table 2)(16). The child with symptoms was given an elimination diet, excluding common allergenic foods. The exact nature of the elimination diet depended on the age of the subject and his or her current dietary intake. Food challenges were conducted on all symptomatic infants, (a) immediately on those who exited the study because of “treatment failure” (vide infra) before age 6 months, (b) between age 7 months and 12 months, and (c) between age 13 months and 60 months.
Data were analyzed using appropriate statistical tests and MINITAB or SPSS or EPI-INFO programs. Prevalence and incidence data were tested by multiple comparison and two-group comparisons using Chi square test with corrections as required. Eczema scores in different feeding groups were not normally distributed and were compared by the nonparametric Kruskal-Wallis test of multiple comparison; for two-group comparisons, the nonparametric Mann-Whitney Rank Sum test was used.
The number of infants who exited and the reasons for leaving the study are shown in Table 3. Between age 0 and 6 months, 42 infants exited the study: six in the partial whey hydrolysate group, 11 in the cow's milk formula group, 11 in the soy formula group, and 14 in the breast-fed group. Some exited because of significant atopic symptoms and signs; such as wide-spread skin rash; wheezing; gastrointestinal symptoms, particularly vomiting; or repeated severe colic. Such an outcome was considered a treatment failure. The number of infants who exited because of atopic symptoms was higher in the cow's milk and soy formula groups, although the differences among the three groups are statistically not significant. Anaphylactic reaction was not observed in any infant.
The cumulative incidence of clinical allergic disease, mainly atopic eczema and asthma, until age 5 years is shown in Table 4. The follow-up until 5 years confirmed and extended the observations at the age of 6 months (9) and 18 months (10). Of the three formula groups, infants fed partial whey hydrolysate from birth until 6 months had the lowest incidence of atopic disease. Differences between incidence of allergic disease in cow's milk formula-fed and whey hydrolysate-fed groups were highly significant. Soy formula feeding was not associated with reduction in allergic disease. Differences between the incidence of allergic manifestations in the whey hydrolysate and breast-fed groups were not significant. Symptomatic infants aged >6 months were treated conventionally, e.g., change of feeding formula, local hydrocortisone cream, inhaled bronchodilators, inhaled corticosteroids, and/or cromolyn.
The cumulative incidence of atopic eczema is shown in Fig. 1. Of the formula-fed infants, those fed partial whey hydrolysate had the least incidence of eczema at all observation points. Soy formula-fed infants did not show any reduction in the occurrence of allergic eczema. The incidence of eczema was comparable in the whey hydrolysate and breast-fed groups. Among affected infants, the severity of skin disease, as estimated by an eczema score, was lowest in the whey hydrolysate group when compared with the breast-fed, cow's milk, and soy formula groups (Fig. 2).
The cumulative incidence of asthma is shown in Fig. 3. The incidence of asthma was lowest in the breast-fed group. Of those fed formula, it was lowest in the partial whey hydrolysate-fed group. There were no significant differences between the incidence of asthma in the cow's milk and soy formula groups.
The period prevalences of eczema and asthma during the ages of 18 months to 60 months (Table 5) were significantly different in the four feeding groups. The lowest prevalance was in the breast and Good Start groups.
The results of skin prick tests and DBPCFC to foods conducted at various ages are shown in Table 6. Foods identified on the basis of positive skin prick tests were chosen for DBPCFC. Those who became symptomatic in the first 6 months of age were tested immediately if the clinical manifestations warranted exit from the study (i.e., treatment failure) or before the introduction of foods other than the assigned formula. Of 24 symptomatic children in the cow's milk group, 16 showed a positive skin prick test to milk, and 12 showed a positive challenge to milk. The corresponding figure for the whey hydrolysate group were 5, 4, and 3, respectively. For the 25 symptomatic babies fed exclusively on soy formula, 19 showed a positive skin test to soy, and 14 had a positive challenge to soy. For ethical reasons, DBPCFC was conducted only in the two breast-fed infants in whom atopic manifestations were severe enough to stop breast-feeding and who exited the study. The differences in incidence of food allergy between the whey hydrolysate group and the cow's milk and soy formula groups are highly significant, but the breast-fed and whey hydrolysate groups did not differ from each other.
When assessed at age 7 months to 12 months after the introduction of other foods, of 28 symptomatic infants in the cow's milk formula group, 17 showed a positive skin prick test to milk proteins. Of these 17 infants, 14 had a positive DBPCFC to milk. Of 12 affected in the whey hydrolysate group, 6 showed a positive skin prick test. Of these 6, 4 were positive on DBPCFC to milk. The corresponding figures for the symptomatic infants in the soy formula group tested for hypersensitivity to soy were 27, 19, and 12, respectively. Results of tests for other foods are shown in Table 5. The differences in incidence of food allergy between the whey hydrolysate group and the cow's milk and soy formula groups are highly significant. Eight breast-fed infants had food allergy. The incidence of food allergy in the breast-fed and whey hydrolysate groups did not differ from each other.
Among those who became symptomatic or continued to show clinical manifestations between ages 13 months and 60 months, 24 in the cow's milk formula group experienced positive food challenges. This number is significantly higher than the 9 symptomatic infants in the whey hydrolysate group. Of 34 symptomatic children in the soy formula group, 19 showed a positive DBPCFC to one or more foods-significantly more than those with food allergy in the whey hydrolysate group.
The results of our randomized double-blind study show that among infants at high risk of atopic disease because of family history, exclusive breast-feeding or feeding a partial whey hydrolysate formula was associated with a lower incidence of allergic manifestations, mainly eczema and asthma, and a lower incidence of food allergy documented by DBPCFC, compared with groups fed conventional cow's milk or soy formulas. The severity of disease among infants with atopic eczema was lowest in the group fed partial whey hydrolysate. The incidence of atopic disorders in the soy and conventional cow's milk groups did not differ. The incidence of allergy in the breast-fed and whey hydrolysate groups was comparable. The follow-up until age 5 years confirmed the persistence of a beneficial effect of breast-feeding or feeding a partial whey hydrolysate formula; thus, these dietary interventions had lessened the likelihood of the development of allergic disease and food allergy.
Some of the important considerations about study design (17) require emphasis. These include stratification of infants into high- and low-risk categories. The possible impact of any prevention program would require a study sample of several thousand infants if only the low-risk infants or the general population of infants is considered. Second, this study was conducted prospectively; in retrospective studies, there is a considerable problem with accurate recall, particularly when the time elapsed has been several years. Even the most motivated mother would find it difficult to recall all events and details of feeding many years later. We felt that it was important to fix sample size before launching the study and to ensure a small drop-out rate. The assessment criteria were objective and quantitative with minimal interobserver error. Because this trial included three different formulas, the assessment was double-blind. Finally, several confounding variables were noted and found to be comparable in the four groups. Apparently negative studies may hide a positive effect simply because of faulty design, such as a small sample size; similarly, false-positive results may be observed when the design is not satisfactory, as in nonblinded trials.
The American Academy of Pediatrics Subcommittee on Nutrition and Allergic Disease (7,8) stated that “to support a claim of allergy prevention, infants from families with history of allergy (biparental or one parent or one sibling) should be fed the product exclusively from birth for at least 6 months under the conditions of a controlled and randomized study and followed for at least one year thereafter (or until 18 months). At that time, they must have a statistically significant lower prevalence of allergy assessed by a clinical scoring system than infants fed a standard milk-based control formula for 6 months and followed until 18 months of age.” The allergic nature of positive clinical manifestations should preferably be supported by double-blind food challenges.
Our observations confirm earlier data (9,10) that indicate the beneficial preventive effect of a partial whey hydrolysate formula when fed from birth in high-risk infants with family history of atopy. The long-term follow-up showing persistence of the beneficial effect until age 5 years suggests that feeding with this hydrolysate is associated with a true prevention of atopy and not just a delay in the development of allergic symptoms. Skin tests showed a lower rate of sensitization in the whey hydrolysate group. Results of radioimmunoassay tests for IgE antibodies were reported earlier (10) and showed a similar trend. The double-blind food challenges confirmed a lowering in the incidence of food allergy in the breast-fed and whey hydrolysate-fed groups, as compared with the cow's milk and soy formula groups.
There is some evidence that an extensively hydrolyzed formula may also prevent atopic eczema (18). However, it must be borne in mind that the extensively hydrolyzed formulas are less palatable and very expensive. Thus, their cost more than offsets the financial benefits of prevention of atopic disease.
The cost-benefit analysis of breast-feeding or feeding various formulas to high-risk infants is shown in Fig. 4. In Newfoundland, as noted, the annual number of births is ≈8,400 per year. About 15% of infants have parental history of atopy, and at least one half of those with parental history show clinical manifestations of allergic disease in the first 5 years of life. Of the symptomatic infants, ≈3% require hospitalization, usually for acute asthma, for an average of 3 days per year. The average cost of treatment of these children with allergic disease until the age of 5 years is ≈$740,000 per annum. Exclusive breast-feeding is an extremely cost-effective strategy for prevention of allergic disease in high-risk infants. In those not breast-fed, feeding a partial whey hydrolysate formula may be expected to lower the cost of treatment to ≈$400,000 per annum. Infants fed an extensive casein hydrolysate from birth also have less allergic disease (18), but the cost of the formula is 3 to 4 times more than that of conventional cow's milk formulas; this high cost makes it economically impractical to recommend such extensively hydrolyzed products for prevention of allergic disease.
Breast-feeding is the preferred mode of feeding infants, including those with parental history of atopy. The benefits of breast-feeding for prevention of allergic disease were confirmed in this study and are enhanced if the mother restricts her intake of common food antigens, such as milk and other dairy products, egg, peanut, fish, and soy (18,19).
If an infant with a family history of atopy is not breast-fed at all, there is overwhelming evidence in favor of choosing a hydrolyzed formula rather than a soy formula. The consensus of opinion shows that soy formulas have no role in prevention of allergic disease (20,21), although these formulas are useful for infants with proven cow's milk allergy (22). There are two main kinds of predigested formulas available, depending on the degree of hydrolysis. The extensively hydrolyzed formulas, based mostly on casein (e.g. Nutramigen, Alimentum) but that could also be derived from whey, soy, or collagen, are the formulas of choice for treatment. However, these products are less well tolerated because of poor taste and are very expensive; the cost is almost three to four times that of conventional formulas. On the other hand, the partial whey hydrolysate formula is effective and well tolerated, and its cost approximates that of conventional formulas. Several studies have shown that feeding partially hydrolyzed formulas to high-risk infants from birth for ≈6 months was associated with a lower incidence of atopic disease, particularly eczema and asthma. The results of these studies have been reviewed (23). In our prospective double-blind study, we found that infants fed exclusively on a partial whey hydrolysate for the first 6 months after birth had a significantly lower cumulative incidence of atopic eczema and asthma until age 18 months; further follow-up described herein shows persistence of the beneficial effect until age 5 years, both in cumulative incidence as well as period prevalence during the ages of 18 months to 60 months, suggesting prevention of allergic disease. Similar results were obtained by Vandenplas et al. (24) and Wilhems et al. (25). Marini et al. (26) confirmed that the incidence of atopic symptoms was lowest among infants in the partial whey hydrolysate group, with follow-up until age 3 years.
In conclusion, effective long-term prevention of atopic eczema and asthma among infants with family history of allergy can be achieved by feeding a partial whey hydrolysate formula for the first 6 months of life. Because the cost of such a formula for prophylaxis is almost comparable to that of standard cow's milk formulas, improving atopic symptoms and food allergy in high-risk infants who are not breast-fed can be accomplished without increasing health care costs, which is an important practical consideration (3).
Acknowledgment: The follow-up study was supported in part by Carnation Nutritional Products, Nutrition Research Education Foundation, and the University Research Professorship Award of the Memorial University of Newfoundland. The author thanks clinical and postdoctoral fellows, laboratory technologists, and nursing staff for their assistance.
1. Taylor B, Wadsworth J, Wadsworth M, Peckam C. Changes in the prevalence of eczema since the 1939-45 war. Lancet
2. Burney PGJ, Chinn S, Rona R. Has the prevalence of asthma increased in children? Br Med J
3. Chandra RK. Food allergy. 1992 and beyond. Nutr Res
4. Chandra RK, Gill B, Kumari S. Food allergy. Ann Allergy
5. Zeiger RS, Heller S, Mellon MH, et al. Effect of combined maternal and infant food allergen avoidance on development of atopy in early infancy. J Allergy
6. Arshad SH, Matthews S, Grant C, Hide DW. Effect of allergen avoidance on development of allergic disorders in infancy. Lancet
7. Subcommittee on Nutrition and Allergic Disease. Infant formulas and allergic disease
. Elks Village, Illinois: American Academy of Pediatrics, 1990.
8. Kleinman RE, Bahna S, Powell GF, Sampson HA. Use of infant formulas in infants with cow milk allergy. A review and recommendations. Pediatr Allergy Immunol
9. Chandra RK, Singh GK, Shridhara B. Effect of feeding whey hydrolysate, soy and conventional cow milk formulas on incidence of atopic disease in high risk infants. Ann Allergy
10. Chandra RK, Hamed A. Cumulative incidence of atopic disorders in high risk infants fed whey hydrolysate, soy and conventional cow milk formulas. Ann Allergy
11. Hanifin JM. Atopic dermatitis. J Am Acad Dermatol
12. Broadbent JB, Sampson HH. Food hypersensitivity and atopic dermatitis. Pediatr Clin North Am
13. Sly RM. Asthma. In: Behrman RE, ed. Nelson Textbook of pediatrics
. Philadelphia: Saunders, 1992:587-96.
14. Ellis EF. Asthma in childhood. J Allergy Clin Immunol
15. Neddenriep D, Schumacher M, Lemen R. Asthma in childhood. Curr Probl Pediatr
16. Bock SA, Sampson HA, Atkins FM, et al. Double-blind placebo-controlled food challenges as an office procedure. J Allergy Clin Immunol
17. Chandra RK. Long-term health consequences of early infant feeding. In: Atkinson SA, Hanson LA, Chandra RK, eds. Breastfeeding, nutrition, infection and infant growth in developed and emerging countries
. St John's Newfoundland: ARTS Biomedical, 1990:47-55.
18. Chandra RK, Puri S, Hamed A. Influence of maternal diet during lactation and use of formula feeds on development of atopic eczema in high risk infants. Br Med J
19. Hattevig G, Kjellaman B, Sigurs N, et al. The effect of maternal avoidance of eggs, cow's milk and fish during lactation allergic manifestations in infants. Clin Allergy
20. Bahna SL. Breast milk and special formulas in prevention of milk allergy. In: Mestecky J, Blair B, Ogra PL, eds. Immunology of milk and the neonate
. New York: Plenum, 1991:445-8.
21. Gruskay FL. Comparison of breast, cow and soy feedings in the prevention of onset of allergic disease. Clin Pediatr
22. Businco L, Bruno G, Giampietro PG, Cantani A. Allergenicity and nutritional adequacy of soy protein formulas. J Pediatr
23. Collin-Williams C. Current status of hydrolysates in the prevention of allergy. In: Chandra RK, ed. Nutrition and immunology
. St. John's Newfoundland: ARTS Biomedical, 1992:379-90.
24. Vandenplas Y, Hauser B, Van den Borre C, Sacre L, Dab J. Effect of a whey hydrolysate prophylaxis of atopic disease. Ann Allergy
25. Wilhems R, Duchateau J, Magrez P, Denis R, Casimir G. Influence of hypoallergenic milk formula on the incidence of early allergic manifestations in infants predisposed to atopic diseases. Ann Allergy
26. Marini A, Agosti M, Motta G, Lusardia C. Dietary prophylaxis in infants with high risk of atopic diseases. Riv Ital Pediatr
ESPGAN Summer School: Pediatric Gastroenterology and Nutrition
August 30-September 6, 1997, Stockholm, Sweden. A one week residential course for young paediatric gastroenterologists at a small conference center by the Baltic Sea near Stockholm with good facilities. The members of the faculty will be present for the entire course. All delegates will present short papers or case presentations.
Faculty: Dr. Henry Asher, Professor Ian W. Booth, Dr. Hans A. Beller, Professor Samy Candranel, Professor J. F. Desjeux, Professor Olivier Goulet, Dr. Alfredo Guarino, Dr. Peter Mila, Dr. Markku Maki, Dr. Joh Puntis, Dr. Rarja Russaka, Mr. Marke Stringer.
Topics: Foregut Motility; Gastroesophagel reflux disease; Gastritits/Ulcer; H. Pylori; pH-monitoring; Motility; The Constipated Child; Upper Endoscopy; Investigations in Paediatric Gastroenterology; Intestinal Transport; Congenital Diarrhea; Research in Intestinal Transport; Coeliac Disease Epidemiology, Clinical Presentation, Investigation, Research; Neonatal Gastroenterology; Neonatal GI-Surgery; Malabsorption; Research in Paediatric Surgery; X-ray and USG; Nutrition Basics; Monitoring Growth; Nutrition General Principles; Neonatal Nutrition; Research in Parenteral Nutrition; Nutrition in ICU; TPN in Neonatalogy; IBD; Surgery in IBD; Histopathology in IBD; Neonatal Gastrosurgery; Short Bowel Syndrome; Research in Short Bowel Syndrome; Team Care of Short Bowel Syndrome; How to Write an Abstract; What Happens to Your Paper When it is Submitted to a Journal; Full Laboratory Day with Western Blot/Northern Blot/PCR/EM microscopy.
Social program: You will not want to miss it!
Number of delegates: Maximum 25
Cost per delegate: Approximately 9.500 SEK (2.000 DM), including accommodations and meals.
For information contact Dr. Yigael Finkel, Paediatric Gasteroenterology, St. Görans Children's Hospital, S-11281 Stockholm, Sweden. Tel: 468-6722148. Fax: 468-6721941.