Health-related quality of life, assessed with a disease-specific questionnaire, in Swedish adults suffering from well-diagnosed food allergy to staple foods

The recruited patients had documented allergy to at least one of the staple foods cow’s milk, hen’s egg or wheat. These particular staple foods were selected since they are difficult to avoid in a Swedish every day diet. The study population consisted of patients with a doctor’s diagnosis of food allergy to any of the three above mentioned staple foods, registered at the outpatient allergy clinic at the Sahlgrenska University Hospital in Gothenburg. The patients were identified and recruited by a clinical dietician during 2010 and 2011 based on medical records. Inclusion criteria were a convincing history of food allergy to at least one of the three staple foods either ascertained by a positive food challenge with objective symptoms, or by high levels of food specific IgE antibodies with strong association to positive double-blind food challenge results according to today´s standard procedure [3, 14‐18]. In total, 103 patients fulfilling the criteria were invited, and received the FAQLQ and EQ-5D questionnaires by regular post together with a formal invitation letter containing information about the study and a pre-paid return envelope. When returning the completed questionnaires an incentive of two movie tickets was offered to the respondents. After one reminder, a total of 80% of the respondents had returned the questionnaires. Two were excluded because the respondents reported not to be allergic to cow’s milk, hen’s egg or wheat any longer, and one person returned a blank questionnaire. Finally, 79 patients (28 males and 51 females) were included in the study (Figure 1). The mean age of the patients was 41 years (range 19–78).

The FAQLQ is available in four versions; an adult form, a teenager form (13–17 years of age), a child form (8–12 years of age), and a parent form for children (0–12 years of age). In this study, we used the adult form, FAQLQ-AF [19]. The validation of the original FAQLQ-AF was done in the Netherlands where it was found to be valid and reliable, and it was able to discriminate between patients with different disease characteristics [20]. The FAQLQ-AF was subsequently translated to eight other languages (including Swedish) and all versions have recently been validated (Goossens et al., personal communication).

The translation of the FAQLQ-AF questionnaire to Swedish was done according to the guidelines set up by the WHO [21]. Thus, the questionnaire was first translated from English to Swedish by a native Swedish speaker (forward translation), then back-translated into English by a native English speaker, and finally the back-translated version was compared to the original English questionnaire. The translations were done by persons with medical skills. A pilot trial was conducted in 10 patients to ascertain that the translated FAQLQ-AF was firmly understood by Swedish-speaking persons. From the results of the pilot study it was concluded that the translation of the instrument to Swedish worked after a few minor linguistic adjustments.

The FAQLQ-AF assesses HRQL in four domains: Allergen Avoidance and Dietary Restrictions (AADR), Emotional Impact (EI), Risk of Accidental Exposure (RAE), and Food Allergy related Health (FAH), containing a total of 29 items (Table 1).

The questionnaire scores are based on a 7-point scale, where 1 is the best possible (highest HRQL) and 7 the worst possible score (lowest HRQL) [10, 20, 22, 23]. Mean HRQL scores are analyzed in each of the four domains and a mean total HRQL score is estimated according to the results in the four domains.

Floor and ceiling effects (percentages of patients with minimal and maximal scores, respectively) of the FAQLQ-AF were investigated in order to verify the validity and reliability of its contents. Such effects were considered to be present if more than 15% of the patients in a sample of at least 50 patients achieved the lowest or highest possible scores, respectively. If floor or ceiling effects are present, it is likely that extreme items are missing in the lower or upper end of the scale. In such cases, as a consequence, patients with the lowest or highest possible scores cannot be distinguished from each other, and reliability of the questionnaire is reduced [24].

Questions regarding the different symptoms that arise as a result of the food intake were included in the FAQLQ-AF. All reported symptoms were aggregated into groups according to the organs that were affected. In the present study, the definition of anaphylaxis includes the four self-reported symptoms “difficulty breathing” and/or “inability to stand”, collapse, loss of consciousness, representing reactions in the respiratory tract and/or cardiovascular system according to definition of anaphylaxis [25‐27]. The classification of symptoms is shown in Table 2.

The generic EQ-5D health-related quality of life questionnaire was used for comparison to the disease-specific FAQLQ-AF. The EQ-5D was developed by the EuroQol Group [28] and comprises of five dimensions (mobility, self-care, usual activities, pain/discomfort and anxiety/depression). Each dimension has three levels: no problems, some problems, severe problems. The questionnaire responses were converted into a single summary index by applying a formula that attaches values (preference-based weights elicited from population surveys) to each of the levels in each dimension. As there is no Swedish index tariff, for this analysis we used the United Kingdom time trade-off value set based on the UK EQ-5D index tariff [29]. An index of 1.0 corresponds to full health (highest HRQL) and 0 the worst possible score (lowest HRQL).

The statistical analysis was performed with the IBM Statistical Package for the Social Sciences (SPSS) Statistics 20. Parametric one-samples T-test was used to test statistical significance with 0.95 percent confidence interval. Linear regression was used to estimate which symptom groups that had the largest negative impact on HRQL.

The study was approved by the Regional Ethical Review Board in Stockholm, Sweden (Dnr 2009/84-31/5), and the collected personal data was treated according to the Swedish personal data act.

Written informed consent was obtained from each patient in this study when responding to the questionnaire for the publication of this report and any accompanying images.

The allergy characteristics of the 79 patients (28 males and 51 females) are shown in Table 3. Briefly, allergies to the staple foods cow’s milk, hen’s egg or wheat were evenly distributed. About half of the patients had a concomitant allergy to tree nuts and/or peanuts. Allergy to shell fish and fish was more common in males, compared to females. About half of the patients had allergy to three or more foods and less than one fifth of the patients had allergy to one or two foods only. Symptoms from the mucous membranes and skin were more common and no differences were seen between men and women, whereas symptoms from the gastrointestinal tract were more common in females. Almost all (90%) of the patients had at least one concomitant allergic disorder (such as asthma, allergic rhino-sinusitis, allergic rhinoconjunctivitis or eczema), and one third were affected by four other allergic disorders in addition to food allergy.

Only few food allergic patients reported the minimal (highest HRQL) or maximal (lowest HRQL) score in each of the domains in the questionnaire. Thus, the analysis showed minimal floor or ceiling effects, which confirms the internal validity of the questionnaire (Table 4).

The scores in the four domains of the FAQLQ-AF are shown in Figure 2. The highest score (lowest HRQL) was found in the domain AADR (i.e. How troublesome do you find it, because of your food allergy, that you …), and the lowest score (highest HRQL) was found in the domain FAH (i.e. How worried are you because of your food allergy …). The mean HRQL score was estimated to 4.85 (95% CI, 4.61-5.10). No significant difference was found in the mean HRQL scores between males and females (4.81 vs. 4.88, respectively).

The presence of concomitant allergic disorders had a profound effect on HRQL in food allergic patients. Patients who, in addition to food allergy, were affected by four other allergic disorders (asthma, allergic rhinitis, allergic rhinoconjunctivitis, eczema and/or other allergic skin problems) reported a lower HRQL compared to those who were affected by three or fewer other allergic disorders (5.21 vs. 4.68; p = 0.03). Moreover, having asthma together with food allergy seemed to have a further profound effect on HRQL compared to patients with food allergy but no asthma (5.03 vs. 4.24; p = 0.01). This is in contrast to the results found in the generic EQ-5D questionnaire, where no difference was seen between patients with and without asthma (mean EQ-5D index value 0.80 vs. 0.79, respectively). Other concomitant allergic diseases did not further affect HRQL negatively, irrespectively of one, two or three concomitant allergic disorders.

The severity of food allergy had an impact on HRQL. Severe food allergy was defined as having a prescription for an epinephrine auto injector (EAI), or self-reported previous episodes of anaphylaxis (i.e. the symptoms “difficulty breathing”, “inability to stand”, collapse and/or loss of consciousness). We found that the total FAQLQ-AF scores were higher in patients who were prescribed an EAI (n = 40), which shows that they had lower HRQL compared to patients who did not have an EAI (n = 39) (5.12 vs. 4.58; p = 0.03). However, such a difference was not found when analyzing the results of the EQ-5D questionnaire (mean EQ-5D index values 0.85 vs. 0.75, respectively). The FAQLQ-AF scores among patients who reported anaphylactic reactions (n = 43) compared to those who did not (n = 36), did not reach statistical significance, although there was a tendency (5.04 vs. 4.63 p = 0.10). When using EQ-5D, no differences were found between patients with and without self-reported anaphylaxis, (mean EQ-5D index values 0.81 vs. 0.79, respectively). The number of offending foods (four or more as compared to three or fewer) did not significantly affect the total FAQLQ-AF score (5.01 vs. 4.67; p = 0.18).

Among the 14 individuals who were allergic to only one staple food, 40% had self-reported symptoms of anaphylaxis. The anaphylaxis diagnosis was subsequently verified from the patients’ medical records. However, there were too few patients for further analyses of each individual allergen. In patients allergic to multiple food items but with no allergy to peanuts and/or tree nuts, 41% reported that they had had anaphylaxis. The corresponding figure for patients with multiple food allergies including allergy to peanuts and/or tree nuts was 62%. The distribution of anaphylaxis and prescription of EAI in the three different groups was as presented in Figure 3. Sixty-five percent of the patients with report of respiratory symptoms and 75% of patients with report of cardiovascular symptoms had an EAI prescription.

The effects of symptoms on the HRQL scores were further analyzed with linear regression. All symptoms were grouped according to the affected organs (Table 3). Symptoms from the skin were reported in 73%, OAS in 63%, mucous membrane in 78%, respiratory in 62%, gastrointestinal in 71% and cardiovascular symptoms in 52% of the patients (Table 3). HRQL was the dependent variable while the symptom groups were independent variables in the regression model. The F-test was significant, indicating a linear relationship between the symptom groups and the HRQL scores and 22% of the variation in the scores could be explained by the symptom groups (R²-value of 0.28; (adjusted R²of 0.22). The regression coefficients were significant for the OAS and cardiovascular symptom groups, however not for the other symptom groups (Table 5). As HRQL decreases with higher points on the scale, the regression analysis demonstrates that OAS and cardiovascular are the symptom groups with most negative effects on HRQL.

Concomitant allergic disorders were common among patients experiencing most of the symptom groups (mean of 2.5-2.8 other disorders). However, patients with respiratory symptoms had only 1.8 other allergic disorders on average. The mean numbers of offending food items were quite similar in the different symptom groups (4.2-5.0 food items) (results not shown).