Does a ‘reverse’ atopic march exist?


Giovanni Passalacqua, MD
Allergy & Respiratory Diseases
Department of Internal Medicine
Padiglione MaraglianoL.go R. Benzi 10
16132 Genoa


The classical description of the atopic march usually refers to the progression from atopic dermatitis towards asthma, but this pathway has been questioned. We assessed in a prospective observation the possible onset of atopic dermatitis in children with asthma alone at baseline, and evaluated retrospectively their characteristics. Seven hundred and forty-five children (360 male, 6–9 years of age) with asthma alone, without food allergy or atopic dermatitis, were followed-up with regular visits for 9 years. 692 children completed the 9-year observation, and 20% of them were found to have developed atopic dermatitis at 9 years. Comparing retrospectively the children who developed AD with the remaining, no significant difference existed at baseline concerning the demographic characteristics and family history. There was a significantly higher proportion ( χ2 = 0.01) of subjects with single sensitization to mites and a significantly lower proportion of polysensitized subjects ( χ2 = 0.01) within the children who developed AD. Sensitization to foods appeared in 9% of children who developed AD and in 3.8% in the other children (NS). According to these observations, the development of a particular allergic disease does not necessarily follow the classical paradigm of the atopic march.

The hypothesis of the ‘atopic march’ in children has gained popularity in the recent years, as it reflects the essentially clinical observation that allergy often progresses from atopic dermatitis (AD) and food allergy, through rhinitis, culminating in bronchial asthma (1, 2). Nonetheless, this hypothesis has also been questioned in its general validity, as the progression is not the same in all children and this progression may lack in some of the stages (3, 4). From a general point of view, it is well known that AD is a risk factor for developing asthma. This is especially true in the ‘extrinsic form’ of AD, which occurs in the context of positive family history, sensitization towards allergens and increased IgE levels. In this case, it is not surprising that children with extrinsic or allergic AD are more prone to develop other allergic disorders such as rhinitis and asthma. What is unclear is whether the chronologic course of the disease should necessarily evolve from AD to asthma or whether different natural history may exist leading, for instance, from bronchial asthma to AD?

In this prospective survey, we aimed at evaluating the clinical course of those children with allergic asthma alone at baseline by following them up for several years (up to nine), in order to assess if they developed other atopic diseases, including AD. After the completion of the prospective study, we assessed in a retrospective way whether some clinical/biological characteristics were specifically related to the progression of the disease.


Children were selected among those referred to our unit between 1993 and 1997.The entry criteria were that they had to suffer only from allergic bronchial asthma (with or without allergic rhinitis) and without the signs or symptoms of AD or food allergy, or clinical history of allergy disorders related to foods. Children, all Caucasian, were included if they presented symptoms like wheezing, prolonged cough or shortness of breath suggesting asthma for at least  1 month prior to entry into the study and significant bronchial reversibililty. The latter was defined as at least a 15% increase in forced expiratory volume in 1 s (FEV1), 20 min after the inhalation of a B2 agonist (5). On the basis of symptoms and lung function tests, the majority of children could be categorized as having intermittent asthma or mild persistent asthma (6). The diagnosis of rhinitis was directly made by pediatricians based on symptoms: sneezing, runny or blocked nose when the child did not have a cold or chest infection. The children had also to have skin positivity to one or more aeroallergens. The aeroallergens included in the skin test panel were those responsible for respiratory allergy in our area: mites, grass, parietaria, mugwort, cat-, dog- and rabbit dander, olive, cypress, alternaria, cladosporium. Food allergens were tested with the prick-by-prick technique, and the panel consisted of milk, egg, wheat, peanuts, fish, hazelnut, peach, apple, tomato and sesame seeds. The lancet was pricked into the fresh material and then immediately into the volar skin of the distal forearm. The test was considered positive if the diameter of the wheal was greater than 3 mm. Histamine hydrochloride 10 mg/ml and saline were used as positive and negative controls. Wheals were measured after 15 min and wheal reactions were outlined with a fine-point marker and transferred to paper with transparent tape. If a patient had a positive result for food(s) then the patient was excluded by the cohort. All the children, once the diagnosis was made, were prescribed an appropriate treatment to achieve the best control of their asthma. In line with the current guidelines, the treatment involved use of inhaled corticosteroids, possibly adding long-acting β2 agonists and/or leukotriene modifiers. For those subjects with mite allergy, avoidance measures were prescribed, including the use of polyvinyl impermeable mattress covers for infants’ cots. Carpets and upholstery were avoided in the infants’ bedroom and lounge. Wherever the control was maintained for at least 4 months a step-down was made. The overall strategy was to calibrate the treatment to the least medication necessary to maintain asthma control. Specific allergen immunotherapy was prescribed when indicated. The children were followed up regularly with scheduled visits once every 4 months or more frequently, according to the clinical need. The families and GPs were instructed to call our unit if either cutaneous signs or pruritus were noticed. AD was diagnosed by a physician of the paediatric department according to the current criteria of SCORAD Index (SI). The SI evaluates the spread of the disease (head and neck, each arm, front and back of the legs, the four trunk quadrants and genitalia) clinical aspects (erythema, oedema, papulation, oozing crusts, excoriation, lichenification, xerosis) and other subjective symptoms (pruritus and sleep loss)(7).


The studied cohort consisted of 745 children, 360 male, aged between 6 and 9 years, with a mean age of 7 years. All of them had bronchial asthma (44% with rhinitis), subdivided into intermittent (57%), mild persistent (38%) and moderate (5%). None had AD at baseline. Of the 745 children, 218 (31.5%) had multiple sensitizations whereas the remaining were mono-sensitized. Fifty-three children were lost to the study during the 9-year observation period: Forty-one did not attend the clinical visits, eight moved to another town, and four developed other chronic diseases (one diabetes, one coeliac disease, two autoimmune thyroiditis). Complete follow-up data at 9 years were therefore available for 692 children (338 males, 354 females, with a mean age of 7 years (range 69) at baseline. Of them, 503 were mono-sensitized (219 for mites, 112 for parietaria, 71 for grass, 26 for olive, 24 for cat, 18 for alternaria, 3 for other).

The first important observation was that, out of 692 children, 141 (20%) had developed AD after 9 years. The percentages of children with AD of new onset were 5.6% at 3 years, 8.5% at 6 years and 5.9 at 9 years. According to the Scoring Atopic Dermatitis (SCORAD) evaluation, within the 141 children, AD was mild in 49 (35%), moderate in 72 (51%) and severe in 20 (14%). The prominent sensitizing allergen in these children was dust mite, positive in 92/141 of the children at baseline (in 60 as single sensitization), followed by parietaria that was positive in 65 subjects (18 mono-sensitized) and grass (26 subjects, 7 mono-sensitized). Of these children, 109 had been treated with drugs alone and 32 with drugs plus specific immunotherapy. Eighty-five (60.3%) patients had intermittent AD, with cutaneous symptoms especially in autumn and/or spring, with long periods of remission. The remaining 56 children (39.7%) had persistent AD and required regular treatment with topical steroids and antihistamines. When compared retrospectively the children who developed AD with those who did not, it could be seen that no significant difference existed at baseline concerning their demographic characteristics (Table 1). Also, there was no difference between the two groups as far as family history was concerned. On the other hand, there was a significantly higher proportion (χ2 = 0.01) of subjects with single sensitization to mites and a significantly lower proportion of poly-sensitized subjects (χ2 = 0.01) within the children who developed AD compared to those who did not (Table 2). In 9 of the 141 children (9%) who developed AD, one or more sensitization to food allergens appeared, whereas this happened in 21/551 (3.8%) of the other children, but this difference did not reach the statistical significance. Of the children who developed AD, 40% had concomitant rhinitis at baseline and this percentage was 48% in the other group.

Table 1.   Demographic and clinical characteristics of the 692 children with allergic asthma who developed (AD+) or who did not developed (AD−) atopic dermatitis
 AD+ (n = 141)AD− (n = 551)
No. of patients141551
Sex (male/female)63/78275/276
Mean age (years)77
Age range (years)6–96–9
Duration of allergic asthma ± SD (years)2.7 ± 1.42.9 ± 1.5
No. of concomitant rhinitis (%) 68 (48%)236 (43%)
Total IgE mean ± SD (kU/l)765 ± 146681 ± 172
Table 2.   Distribution of the sensitizations in children who developed or who did not developed AD
AllergenAD+, n (%)AD−, n (%)
  1. *Chi-square P < 0.05.

HDM60 (42)*85 (15.4)
Parietaria18 (13)37 (6.7)
Grass7 (5.5)26 (4.8)
Olive tree14 (2.6)
Cupressus11 (2)
Artemisia4 (0.7)
Alternaria2 (1.5)19 (3.4)
Cladosporium9 (1.7)
Cat dander18 (3.4)
Dog dander10 (1.8)
Rabbit dander3 (0.5)
Multiple sensitizations54 (38)*315 (57)
Total141 (100)551 (100)


The existence of an ‘atopic march’ leading from AD to asthma is nowadays a matter of debate, as the classic progression of allergic diseases in children is not all the same (8). Certainly, it is well known that AD in the early infancy (9) is a risk factor for the development of asthma in the subsequent years, in similar way it happens with rhinitis that often precedes the onset of asthma (10). Nonetheless, if we consider allergy as a systemic disease with different clinical manifestations, we can also suppose that the existence of one of the clinical diseases (rhinitis, asthma, AD) is only the emerging signal of the presence of the atopic status and that any other disease can appear during the natural history, as happens in subjects with asthma alone who develop rhinitis later (11). Our data would support this hypothesis as 20% of children with asthma only at baseline developed AD later in their life. Surprisingly, the only detectable difference between those children who developed AD and who did not was the single sensitization to dust mite, whereas the positive family history for atopy was not discriminative. The study has possible limitations, especially as part of the assessment was retrospective. On the other hand, the population was well selected at baseline, confounding factors (e.g. food allergy) were excluded, and the diagnosis of AD was carefully made by a paeditrician. Another criticism maybe that the study is only clinical, but in this sense the atopic march is essentially a clinical concept, based on the presence of certain diseases, and can therefore be studied only at a clinical level. Our observations raise the question if asthma can be considered paradoxically a risk factor for the later development of atopic dermatitis.