Some selected characteristics of the collective are summarized in Table 2 (demographic, general medical data and environmental exposure), Table 3 (specifics of work-related symptoms) and Table 4 (diagnostic evaluation), a more comprehensive overview of the personal and medical characteristics and the working conditions is presented as an additional online file (see Additional file 3). Aspects that need to be further addressed are followed up in the text with additional information.

Concerning private pet contacts (reported by 39%), cats (n = 26) and dogs (n = 30) were the most commonly named pet types (multiple answers possible). From a medical standpoint a positive reaction in the SPT to the own pet type might be of further importance and was present in 19% of the cases with private pet contacts (corresponding to 7% overall).

Besides pets as important extra-professional exposures other ubiquitous allergens might influence airway symptoms and diseases as well. In this context, sensitization to environmental inhalation allergens is also relevant. From the SPT performed during consultation and other external findings such sensitization could be assumed in 63% of the collective. Grass/grains and tree pollens were the most common positive allergens. Based on the reports of a known allergic rhinoconjunctivitis or asthma to environmental allergens, anamnesis suggested that at least 66% of those sensitizations were clinically relevant. However, the shortened atopic diathesis score overall was indicative of a rather low atopic diathesis prevalence.

Regarding the specifics of the work-related anamnesis the following aspects shall be further pointed out. While representing the full spectrum of the working population, tenure time as a hairdresser and age at initial symptoms were not normally distributed but rather clearly left-skewed respectively suggestive of a bimodal age distribution (Fig. 1). Many cases had also described an expansion of initial work-related symptoms over time (Fig. 2). Especially a progression from pure upper to additional lower airway symptoms (as seen in 17%) can be of high medical relevance. Only very seldom no progress towards the lower airways occurred. While most cases named hairdressing specific triggers (certain substances or stay in the saloon in general) for their airway symptoms, two cases described no hairdressing specific exposures as triggers at all, but rather an aggravation by work stress and smell of customers’ perfumes. Usually workplace symptoms started within 10 min of exposure (73% of cases with available data (n = 79)). When a seasonal fluctuation of symptoms was reported, such aggravation was attributed to weather-conditions (warm-humid, cool-damp), reduced ventilation in winter or seasonal allergens.

Tenure as a hairdresser (n = 137; grey) and age (n = 145; black) at initial symptoms at work

Expansion of indicated initial symptoms (dotted line) during course of the disease (black line)

Besides airway symptoms some cases also described an immediate reaction in form of a contact urticaria to hair/blonde dyes or AP, mostly concomitant with the first airway symptoms. Other skin alterations were described as well, especially hand eczema and known contact sensitizations to hairdressers’ substances. From those cases with a contact sensitization against AP, three cases described also an urticarial reaction to blonde dyes.

With respect to the diagnostics and medical evaluation it seems important to point out, that despite the synopsis of current lung function data (bodyplethysmography and MCT) from the consultation only allowed the diagnosis of an OVD in 49%, overall an obstructive lower airway disease could be confirmed in 60%. For this, accountable previous lung function results were also considered for the final diagnosis, when available, to account for possibly concealing effects like premedication (the antiobstructive medication active during lung function testing included bronchodilatators in 34%) or exposure abstention.

For evaluation of a specific immediate type reaction to workplace substances SPT were performed where possible (2x contradictions: pregnancy and severe previous anaphylactic reaction not suitable for ambulant testing; 1x patient refusal). Almost all cases were tested with AP or a test solution of their own blonde dye (in one case the reason for presentation was not the former hairdressing job), but not all tests produced appraisable results due to insufficient controls. Furthermore, henna was tested in a smaller subset due to exposure anamnesis. Previous skin tests to AP or blonde dye from other institutions/consultations were considered as well. Overall, positive or questionable skin reactions to AP or blonde dye were present in n = 37 cases (one with positive reaction in SPT on-site and previous positive test). To evaluate the clinical relevance of those skin reactions, exposure-specific anamnesis and additional SIC were also considered. By this, clinical relevance of the skin test reactions to AP or blonde dyes could be established for n = 22 cases (15%). Furthermore, two cases without appraisable SPT to AP (due to urticaria factitia) had a positive SIC (for more details see case-by-case breakdown in additional file 2). Therefore, a specific occupational causation for an obstructive airway disease was overall confirmed in 16% (n = 24), although in the two last-named cases this could not be surely attributable to a specific hypersensitivity reaction to hairdressers' substances (invalid SPT).

For henna SPT gave two positive reactions. One without clinical relevance and one indicative of a clinically relevant specific hypersensitivity (contact urticaria to previous henna tattoo and in unprotected skin contact to clients’ self-henna-dyed hair), however it could not be considered the primary causative agent for the work-related respiratory symptoms as no henna was utilised in the saloon directly (see also additional file 2).

Latex was also pricked in some cases and gave five positive or questionable SPT reactions but was not rated as relevant for the respiratory symptoms in any case due to anamnesis.

Figure 3 shows the overall distribution of case categories among the collective. While in about 45% a specific occupational causation from hairdressing was unlikely, it was unclear in 17% and likely or proven in almost a quarter of the collective. Due to no objectifiable respiratory diagnosis no occupational causation could be discussed for 14%.

Case classification according to certainty of a specific occupational causation of obstructive airway disease

Among group 1 and 2 (confirmed or likely causation) cases with a rhinitis were slightly more common than cases with an obstructive lower airway disease (= widely referred to as asthma) as shown in Table 1. However, even more common was a combination of both affected systems (lower and upper airways). Applying the most severe criteria an occupational asthma was asserted in the 15 cases with an obstructive lower airway disease and confirmed occupational causation (10%) and an occupational rhinitis was asserted for 13% (n = 19 cases), but with some significant overlap between both disease entities (for details see Table 1). Comparison of patients’ characteristics among the defined categories 1 to 5 for substance-specific causation revealed several interesting aspects (full breakdown for the five categories included in the supplementary table in Additional file 3). For additional visual illustration of the magnitude of subgroup variations an orientating heat map of selected parameters is also included in the Additional file 4.

While age and tenure time at initial symptoms are naturally associated (as shown by a supplementary figure, see Additional file 5), both were considerably lower in group 1 than in groups 2 to 5 (see supplementary table in Additional file 3). Analysis of variance confirmed a significant difference among groups (age (n = 144): H = 17.182, p = 0.002; tenure (n = 136): H = 15.387, p = 0.004). Post-hoc tests revealed group 1 vs. 4 for age (p < 0.001) and tenure (p = 0.001) and group 1 vs. 5 (p = 0.046) for age as differing. Accordingly, the bimodal distribution of age and left-skewed tenure time among the overall collective (Fig. 1) is largely an effect based on group 1, as shown by group-wise depiction of those parameters (as shown by a supplementary Fig. S2, see Additional file 6).

Contrary smoking behaviour was comparably distributed among groups and variance analysis for pack years (n = 51 data) also showed no differences (H = 7.261, p = 0.123).

Interestingly, group 5 not only missed an OVD diagnosis and rhinoconjunctival symptoms (by definition) but also showed fewer hand eczema. In contrast, group 1 and 2 had an increased prevalence of known contact sensitization against hairdressers’ substances and especially AP. Furthermore, they more often complained of urticaria to blonde dyes and AP.

Also features concerning atopy showed significant differences among groups. Group 1 showed a higher total IgE average and more cases with increased IgE (> 100 U/mL) while the opposite was true for groups 4 and 5. Variance analysis confirmed significant group differences for IgE (H = 17.209, p = 0.002; n = 142) with the post-hoc tests showing group 1 differing from group 4 (p = 0.015) and 5 (p = 0.038). Accordingly, group 1 showed the highest average SADS and group 5 the lowest. Here post-hoc analysis of the significant group variances (H = 14.416, p = 0.006; n = 147) pointed to group 5 as the differing one (group 5 vs. 1: p = 0.003; group 5 vs. 3: p = 0.042; group 5 vs. 4: p = 0.035). This was further underlined by the high percentage of cases with 0 points in the SADS for group 5. Furthermore, the prevalence of allergic rhinoconjunctivitis, positive SPT and type-I-sensitization to environmental inhalation allergens was much less common in group 5, but tended to be increased in group 1 and to lesser extent group 3.

However, some subgroup abnormalities, especially concerning nature of workplace symptoms, lung function data, AP SPT and final diagnoses, were primarily caused by category definition (see Table 1).