European Respiratory Society


The effect of dietary factors on asthma is controversial. This study examined food consumption and the use of fats in relation to wheezing and allergic rhinitis in children.

Baseline questionnaire data on individual and family characteristics were recorded by parents of 5,257 children aged 6–7 yrs living in central Italy participating in the International Study on Asthma and Allergies in Childhood study. A total of 4,104 children (78.1%) were reinvestigated after 1 yr using a second parental questionnaire to record occurrence of respiratory symptoms over the intervening 12 months. Consumption of foods rich in antioxidants, such as vitamins C and E, animal fats, and food containing omega‐3 fatty acids were investigated using a food-frequency questionnaire. Frequency of use of fats was also evaluated. Wheezing, shortness of breath with wheeze, and symptoms of allergic rhinitis in the past 12 months were considered.

Intake of cooked vegetables, tomatoes, and fruit were protective factors for any wheeze in the last 12 months and shortness of breath with wheeze. Consumption of citrus fruit had a protective role for shortness of breath with wheeze. Consumption of bread and margarine was associated with an increased risk of wheeze, while bread and butter was associated with shortness of breath with wheeze.

Dietary antioxidants in vegetables may reduce wheezing symptoms in childhood, whereas both butter and margarine may increase the occurrence of such symptoms.

An increase in the prevalence of asthma over the last decades has been seen in developed countries 14. The marked changes in the western diet, with decreased consumption of fresh fruit, vegetables, fish and milk, and the increased intake of foods rich in fats, have led to the hypothesis that the epidemiological changes of asthma are associated with the changes in food consumption 57. Three main hypotheses on the effect of diet on asthma have been formulated, including excess in sodium intake 5, deficiency in dietary antioxidants 8, and changes in dietary balance between omega‐3 polyunsaturated fatty acid (PUFA) (fish oil, fish and shellfish, and leafy vegetables), and omega‐6 PUFA (vegetable fat, such as margarine and processed foods) 9. The sodium hypothesis has been well studied and its impact on the increase in asthma incidence remains controversial 10, whereas antioxidants and fatty acids are appealing, since they seem to influence the immunological and inflammatory responses that are of key importance in asthma 1114. Lack of beta-carotene, vitamin C and vitamin E may increase susceptibility to oxidant attack and airway inflammation 12, while dietary fats may have negative effects on immune and inflammatory responses 13. More recently, other nutrients have come into focus, such as magnesium intake 15, saturated fats 13, and other antioxidants, like selenium and consumption of flavonoid-rich foods and drinks 16.

Few studies have investigated the role of diet on childhood asthma, and the available investigations, with only one exception 17, have considered only individual food groups 14, 1821. To evaluate the association of several dietary factors to wheeze and allergic rhinitis, the current study examined data collected in a large study performed on children living in two areas of central Italy.


The Italian Studies on Respiratory Disorders in Children and the Environment (SIDRIA) study was conducted in Italy between October 1994 and March 1995, as a part of the International Study on Asthma and Allergies in Childhood (ISAAC) initiative 1. The survey was designed to estimate the prevalence of asthma and other atopic diseases in children and to investigate the potential risk factors for asthma. The study was conducted among children aged 6–7 yrs attending primary school, in eight centres of northern and central Italy. A detailed description of the methods of the study has been reported elsewhere 4, 22. The parents had to complete the Italian version of the ISAAC questionnaires at home. It included questions on a number of risk factors for childhood asthma and other respiratory and allergic diseases. Socio-demographical variables, such as parental education, household crowding, maternal and paternal smoking, presence of dampness and/or mould in the child's bedroom, and parental asthma, were collected.

A follow-up study, 1 yr after the first data collection, was conducted in two of the participating centres (Rome and Viterbo, a small town ∼80 km from Rome); the methods have already been described by the authors in a previous paper on fresh fruit consumption and wheezing 20. A new questionnaire was sent to the parents of 5,257 children to record information on respiratory symptoms, including shortness of breath with wheezing, over the previous 12 months. In addition, parents filled in a detailed semiquantitative food frequency questionnaire. A total of 18 food items were investigated to study foods rich in antioxidants, such as carotenoids and vitamins C and E (cooked vegetables, salads, tomatoes, peppers, potatoes, fresh fruit, citrus fruit, kiwi, nuts), animal fats (bread and butter, milk, cheese, liver), vegetable fats (margarine, olives), and food containing omega‐3 fatty acids (blue fish). The questionnaire assessed weekly consumption using a five-level scale: never, less than once a week, 1–2 times per week, 3–4 times per week, 5–7 times per week. Fresh tomatoes are important sources of vitamins and antioxidants but production and consumption are different by season. The loss of vitamins that occurs in storage and processing during the winter 23, 24, led the authors to study the consumption of both in season (summer) and out of season tomatoes. The questionnaire also contained questions about the frequency of use of fats for cooking, dressing, and sauces, asking whether butter, margarine, olive oil, and seeds oil were used never, sometimes, often, or always. Sauces with butter or margarine in Italy are usually eaten with pasta or rice.

In the present analysis, any wheezing in the past 12 months, the occurrence of attacks of shortness of breath with wheeze in the past 12 months, and the occurrence of symptoms related to allergic rhinitis (sneezing, or a runny/blocked nose apart from common cold/flu with itchy watery eyes) were the three outcome variables to be evaluated in relation to diet. Other respiratory symptoms (nocturnal dry cough and chronic cough apart from colds) were also evaluated, but for the sake of simplicity the authors decided to report in detail only those symptoms chosen a priori. Logistic regression was performed to analyse the association between food consumption (in five levels, using “never” as reference) or fats used for cooking, dressings and sauces (“never” was the reference) and the outcome variables, after adjusting for the following confounders: sex, study area, paternal education, household crowding, maternal smoking, paternal smoking, dampness or mould in the child's room, and parental asthma. When a specified category of food consumption had only few subjects (∼≤1%), it was combined with the category of lower consumption. Odds Ratio (OR) and 95% confidence interval (95% CI) were computed to estimate the degree of association. Moreover, food consumption was included as an ordinal variable in the logistic model to test for linear trend (Wald test). Spearman's correlation coefficients were computed to evaluate the degree of correlation among dietary variables.


The response rate to the study was 78.1% for a total of 4,104 children. The response rate was higher in Viterbo (85.9%) than in Rome (75.7%) and it did not vary by sex, rate of parental asthma, or father's education. The characteristics of the children who participated in the study are shown in table 1. Among the 4,104 children, almost 75% lived in Rome, 37% were exposed to maternal smoking and 46% were exposed to paternal smoking. Occurrence of any wheeze, shortness of breath with wheeze, and rhinitis symptoms in the past 12 months were 5.8%, 3.8% and 8.5%, respectively.

Tables 2 and 3 present the weekly consumption of the 18 different food items and the use of fats for cooking, dressing, and sauces. The last column indicates the number of subjects for which information was available. A relatively frequent consumption of milk, tomatoes, and fresh fruit (in particular citrus fruit) and an infrequent use of animal fats (bread and butter, and bread and margarine, and cheese) were recorded. The rare use of bread and butter and bread and margarine is not a surprise given the traditional diet in central Italy. Olive oil was the principal fat used for cooking and dressing. When the matrix of Spearman's correlation coefficients was examined, there was a moderate correlation (maximum 0.50) among foods pertaining to the same food group (orange/kiwi, salad/tomatoes), while the correlation among the dietary variables pertaining to different food groups was low, with the highest values found for fruit and vegetables (0.30). The correlation between consumption of bread and butter and bread and margarine was also low (0.23).

Tables 4–9 report the adjusted OR and 95% CI of the association of wheeze, shortness of breath with wheeze, andallergic rhinitis during the last 12 months with thefrequency of various foods intake and uses of fats. The p‐value for trend evaluates the potential linear relationship with the outcome variables across the categories of food consumption.

A protective effect of fresh vegetables and fruit on wheeze (during the past 12 months) was observed (table 4). In particular, reduced wheezing corresponds to a higher intake of salads (p=0.047). The effect is even stronger for those consuming summer tomatoes; the OR for children who eat tomatoes more than four times per week in comparison with those who do not was 0.49 (95% CI: 0.31–0.80, p=0.003). A protective effect was also seen in children who frequently consume nuts (p=0.017). OR below one and indications of a trend were seen in the frequent intake of fresh fruit, citrus fruit, and kiwi.

A significantly higher prevalence of wheezing was noted among children who eat bread and margarine at least once per week in comparison with children who never eat bread this way (OR=2.52; 95% CI: 1.25–5.09) (table 5). Butter used as a cooking fat was positively associated with wheeze (p=0.031). Moreover, those who usually use butter on pasta had an increased risk for wheeze (OR=2.85; 95% CI: 1.01–7.42).

Confirming the results for “any wheeze”, summer tomatoes and fresh fruit, in particular citrus fruit, had a beneficial effect in preventing shortness of breath with wheeze (table 6). The OR of children who consume tomatoes, fresh fruit, and citrus fruit more than four times per week in comparison with those who do not eat these foods were 0.55 (95% CI: 0.32–0.96), 0.37 (95% CI: 0.16–0.85) and 0.59 (95% CI: 0.35–1.00), respectively and p=0.031, 0.022, and 0.016, respectively. Bread and butter consumed more than four times per week (OR=3.12; 95% CI: 1.18–8.23) was positively associated with shortness of breath with wheeze.

Only three associations were noteworthy for allergic rhinitis: consumption of nuts, milk intake, and butter in sauces (tables 8 and 9). Children with frequent consumption of nuts (more than four times per week) had a higher risk of rhinitis (OR=2.12, 95% CI=1.21–3.71). A significantly lower prevalence of nasal symptoms was noted among children who drink milk more than three times per week in comparison with children who never drank milk (p=0.011). Butter used for sauces was positively associated with frequency of rhinitis (p=0.047).

Table 10 summarises the main results of the associations between fruit and vegetables consumption and fats consumption with the two wheezing symptoms and also illustrates the findings for the same food items in relation with nocturnal dry cough and chronic cough. The OR of the comparisons between the highest versus the lowest consumption categories are reported, together with the results of the trend test across the consumption categories. The protective effects of cooked vegetables, summer tomatoes, fresh fruit, and citrus fruit were consistent for the four respiratory symptoms investigated. The negative effects of butter (with bread or for cooking) were present only for wheezing symptoms and not for cough symptoms whereas margarine (with bread) was more or less associated with all the four symptoms.


The current study found a protective effect for vegetable and fresh fruit consumption and a negative effect for butter and margarine use on wheezing symptoms in children. Apart from butter, none of these foods were associated with allergic rhinitis. No effect for fish containing omega‐3 fatty acids was found.

The present analysis confirms the protective role of citrus fruit on wheezing 20, and suggests associations with other dietary items. The protective effects seem to be clear for vegetables, in particular tomatoes, and fresh fruit. Both “any wheeze” and “shortness of breath with wheeze” decreased in relation to an increased consumption of summer tomatoes and fresh fruit. Vegetables and fruit are the major source of antioxidant vitamins, such as vitamin C and carotenoids. Dietary intake of these vitamins may exert an activity against oxidative lung damage and may decrease the inflammation of the airway 10. It has been suggested that vitamin C is the major antioxidant substance in the airways and it protects against endogenous and exogenous oxidant molecules 25. Vitamin C is the most abundant antioxidant substance in the extracellular fluid in the lung and contributes to the regeneration of membrane-bound oxidised vitamin E, allowing it to function again 26. Vitamin E, of which olive oil is the principle source in Mediterranean diet, is a fat-soluble constituent of the cell membrane and represents the body's principal defence against oxidant-induced membrane injury, breaking the lipid peroxidation chain reaction. Vitamin E seems to suppress neutrophil migration 13, and inhibits immunoglobulin (Ig)E production 14.

Some evidence on the role of vegetables and fresh fruit on asthma-like symptoms have been found in children. Apart from the authors previous results 20, in a cross-sectional study in England and Wales 19, children who never ate fresh fruit had an forced expiratory volume in one second of 79 mL (4.3%) lower than that of children who ate fruit more than once a day, and this association was even stronger in children with wheezing. A case-control study on childhood asthma in Saudi Arabia found a strong protective effect for vegetable consumption and Vitamin E intake 17. An ecological analysis of the ISAAC study involving 53 countries (children aged 13–14 yrs) has shown a consistent pattern of a decrease in wheezing symptoms, allergic rhinoconjunctivitis, and atopic eczema associated with an increased per capita consumption of vegetables, vegetable nutrients, and nuts 27. A recent study involving 20,000 children from central-eastern European countries has found a protective effect of fruit and vegetables consumption on the occurrence of cough and on wheezing symptoms and the effects were stronger when considering summer fruit and vegetables 28. In adults, an increase in bronchial reactivity associated with lower intake of vitamin C was found 29, while Bodner et al. 30 found that a higher intake of vitamin E was related to a lower risk of asthma onset. The current study shows a protective effect for nuts on the prevalence of shortness of breath with wheeze, and a similar result was found in the Nurses' Health Study. Women participating in that study had lower incidence of asthma in the highest quintile of vitamin E intake, and the effect was attenuated after excluding nut consumption 31. However, nuts not only contain vitamin E but also magnesium, which has been shown to protect against asthma among adults 15, though it should be noted that the association with nuts is based on relatively few subjects and might just be a spurious finding.

It is difficult to single out the most important antioxidant nutrient responsible for the protection, since the amount of vitamins C and E, and carotenoids are highly correlated. However, of note is that among the types of vegetables and fruit investigated, the strongest result was the protective role of summer tomatoes. In the longitudinal Nurses' Health Study 31, consumption of tomato juice and tomato sauce was inversely related to asthma incidence. Moreover, a study regarding Sudanese children has linked tomato consumption with a reduction of respiratory infections 32. Tomatoes contain lycopene, a carotenoid with a strong antioxidant property, and other potentially beneficial substances, including vitamin C and E 33. High intake of tomatoes or lycopene has been associated with decreased cancer risk 34, in particular lung, stomach and prostate cancer. The results from the current study are a clear indication that more research is warranted on the potential beneficial effects of foods containing lycopene.

The current study shows a harmful effect of both margarine and butter consumption in relation to “any wheeze” and “shortness of breath with wheeze”. Butter used for pasta was also correlated with allergic rhinitis. Margarine is rich in polyunsatured fatty acids, particularly omega‐6 fatty acid, while butter is an animal fat rich in saturated fatty acids. The Zutphen study revealed that high intake of PUFA and linoleic acid was associated with high incidence of chronic lung disease 35. A recent ecological study regarding 37 centres in Europe showed a positive association between daily intake of monounsaturated (MUFA) fatty acid and prevalence of allergic sensitisation in adults 36, while the Nurses Health Study showed inverse association between the intake of MUFA and linoleic acid and asthma incidence 31. Among preschool Australian children, the risk of asthma was higher for a high consumption of PUFA 21. A cross-sectional study conducted in Germany found an association between margarine consumption, compared to butter use, and rhinitis symptoms. 37. A possible explanation for these findings is that high intake of fats leads to an increased ratio of omega‐6 to omega‐3 fatty acids. In general, PUFA have been shown to modify cell-mediated immunity, as well as IgE production and allergic responses 38, but omega‐3 PUFA and omega‐6 PUFA have different effects on inflammation. The omega‐3 PUFA have a specific role in controlling inflammation, but are easily replaced by omega‐6 PUFA, which exert the opposite activity. Atopic patients have an altered composition of PUFA in serum and cell membranes, while a balance between n‐3 PUFA and n‐6 PUFA metabolism exists in nonallergic children 39. However, a complete explanation for the biological mechanism that involves PUFA on airway inflammation has not been found 40.

While PUFA have been relatively well studied, less attention has been paid to the role of saturated fats in inflammation. It has been suggested that saturated fats may modify serum cholesterol levels and the cell membrane content of arachidonic acid, both of which may affect lymphocyte function. However, how saturated fats may modify airway inflammation remains unknown 13. From an epidemiological perspective, there are two observations on the role of saturated fats on asthma. Results from the First Nutrition and Health survey in Taiwan suggested an increase of the risk of asthma in teenagers corresponds to an increased intake of saturated fats, while MUFA fats were inversely related with asthma 13. The case-control study of adults by Bodner et al. 30 found a protective effect of low saturated fat consumption.

The biological mechanism underlying the association between rhinitis symptoms and milk intake is difficult to explain. Although milk consumption has been considered as a proxy for vitamin A intake and lower prevalence of chronic bronchitis has been observed among subjects drinking milk daily versus never 41, a bias could have occurred. Symptomatic children may avoid milk because of food intolerance, since milk is one of the most allergenic foods 25, and the observed effect may be due only to such selective avoidance. However, the association between allergic rhinitis and nuts consumption is probably correct, since the allergenic properties of nuts are well known 25. It is worthy of note that, like in the study conducted in Taiwan on the role of diet on asthma and allergic rhinitis 13, apart from butter used on pasta, none of the dietary factors that were associated withwheeze showed a significant association with allergic rhinitis. This finding suggests that dietary components act via an inflammatory mechanism, other than eosinophilic inflammation, to produce increased bronchial reactivity and reversible airways obstruction.

Several limitations of the study should be noted and caution needs to be taken in interpreting the results due to the following factors: 1) The authors have considered parental reports of children's diet and symptoms, and information bias could have occurred. Although questionnaire data on dietary items recorded by parents are considered to be accurate enough, especially regarding fruit and vegetable intake 42, the questionnaire was not validated against a standard. It is well known that brief food frequency questionnaires are useful to analyse the association between nutrient intake and chronic disease, but they are subject to substantial measurement error and dietary change 43; 2) The analysis considers symptoms and diet collected at the same time; 3) Although the authors considered more foods and food groups than other studies, the scope of the nutrients that were investigated is limited and information of total energy intake is unavailable; 4) It is difficult to single out the specific factor when food patterns, rather than a specific food, may be associated with the outcomes. However, the authors did not find a high correlation among the dietary variables that were investigated; 4) When investigating protective effects, there is always the possibility of “reverse bias”, namely that those with symptoms purposely avoid these foods; 5) Some analyses are based on few subjects and the results may be due to chance; 6) The authors do not have information on height and weight, possibly relevant variables for wheezing.

In conclusion, the current study has suggested a preventive role of a diet rich in vegetables and fruit and lowinfatty foods on childhood wheeze. However, the roleof dietin the development of childhood asthma, the relevant time windows (including during pregnancy) and induction time, and the biological mechanisms underlying the associations, all need to be further investigated before promoting aspecific diet for primary and secondary asthma prevention.

View this table:
Table 1

Characteristics of the study subjects

View this table:
Table 2

Description of food consumption

View this table:
Table 3

Description of use of fats in cooking, dressing, and sauces

View this table:
Table 4

Association between food consumption and 12 month occurrence of wheeze

View this table:
Table 5

Association between use of fats in cooking, dressing and sauces and 12 month occurrence of wheeze

View this table:
Table 6

Association between food consumption and 12 month occurrence of shortness of breath with wheeze

View this table:
Table 7

Association between use of fats in cooking, dressing and sauces and 12 month occurrence of shortness of breath and wheeze

View this table:
Table 8

Association between food consumption and 12 month occurrence of allergic rhinitis

View this table:
Table 9

Association between use of fats in cooking, dressing and sauces and 12 month occurrence of allergic rhinitis

View this table:
Table 10

Summary results of the association between fruit and vegetables and fats with respiratory symptoms


The authors would like to thank P. Compagnucci and M. Huber for their help in editing the paper.


  • For editorial comments see page 719.

  • Received January 17, 2003.
  • Accepted May 28, 2003.


View Abstract