Hand eczema among hairdressing apprentices in Denmark following a nationwide prospective intervention programme: 6-year follow-up

Contact Dermatitis Original Article COD Contact Dermatitis Hand eczema among hairdressing apprentices in Denmark following a nationwide prospective intervention programme: 6-year follow-up Sanne S. Steengaard, Anne Bregnhøj and Jeanne D. Johansen Research Centre for Hairdressers and Beauticians, Department of Dermato-Allergology, University Hospital Gentofte, 2900 Hellerup, Denmark doi:10.1111/cod.12588 Summary Background. Hand eczema is the commonest occupational skin disease in Denmark, and hairdressing is a high-risk profession. In 2008 2010, a clinically controlled, prospective intervention study aimed at reducing the development of hand eczema was conducted at hairdressing schools in Denmark. The findings showed that significantly fewer apprentices in the intervention group developed hand eczema over a period of 18 months. Objectives. To investigate the long-term effect of the intervention. Methods. Two hundred and eighty-four participants were identified from the original dataset, and were sent a questionnaire. Results. No difference was seen between the intervention and control groups. This may partly be attributable to the two groups no longer being well matched, and improved work habits in the control group. Overall, there was an improvement in work habits. Participants had a 1-year prevalence of hand eczema of 22.4%. Reactions to hair dye were reported for 24.5%, and 35.5% had left the trade; 36.4% used gloves when shampooing, and 21.3% stated that they cut hair before colouring it. Conclusions. The effect of the intervention was not visible after 6 years, but an overall improvement in work habits was noted. Key words: follow-up; glove use; hairdressers; hair dye exposure; hand eczema; intervention study; occupational contact dermatitis; quality of life; work routines. Hand eczema is the commonest occupational skin disease in Denmark, and hairdressing is a profession with a high incidence of occupational contact dermatitis (1, 2). Hairdressers and hairdressing apprentices are at increased risk, owing to their excessive contact with allergens and irritants, combined with wet work (3 7). Hairdressers in Denmark remain in the profession for an average of 8.4 years, including their training period. Hand eczema is a frequent reason for leaving the trade (8). Hairdressing apprentices in Denmark have a prevalence of hand eczema of 34.5% during their 4-year Correspondence: Sanne Skovvang Steengaard. National Allergy research Center, Gentofte University Hospital, opg 20A. DK-2900 Hellerup. Tel: +45 38 67 73 09 E-mail: sanne.skovvang.steengaard@regionh.dk Accepted for publication 7 March 2016 training (9), which correlates well with a prevalence of hand eczema of 27 58% reported in previous studies (10, 11). Evidence-based training of workers in high-risk occupations has shown good results in preventing hand eczema for periods of 3 12 months (12, 13); however, the long-term effect is unknown. In 2008 2010, a study was conducted at hairdressing schools in Denmark, aimed at reducing the development of hand eczema (7). The study had a clinically controlled, prospective intervention design, and included all state-approved hairdressing schools in Denmark. It was shown that hairdressing apprentices who participated in the intervention used gloves during wet work more often, and, in general, spent less time with wet hands than those in the control group. In the intervention group, significantly fewer participnats developed hand eczema over a period of 18 months. Of the participants who had hand 32 Contact Dermatitis, 75, 32 40

eczema during the study, those from the intervention group more often consulted their general practitioner and a dermatologist, and only those from the intervention group reported their condition to the Board of Occupational Health. Overall, the study showed that the intervention had reduced the incidence of hand eczema, and increased the awareness of hand eczema and protective work habits among the apprentices. Following the positive results from the intervention, the educational programme from the intervention schools was implemented in 2011 at all hairdressing schools in Denmark. The aim of the present study was to investigate the long-term effect of the intervention. Materials and Methods Primary investigation, 2008 2010 In the 2008 2010 investigation, the schools were divided in two groups: four intervention schools and six control schools. At the intervention schools, selected teachers underwent a 2-day course in general skin physiology and prevention of hand eczema and allergy. Together with the primary investigator, the selected teachers developed an educational programme for the intervention schools. The control schools continued with the traditional training, and no teachers were given extra education. All apprentices were enrolled within the first 2 weeks of their education. The study had three data collection points: inclusion, also called baseline; first follow-up after 6 months; and second follow-up after 18 months. On each occasion, the hairdressing apprentices completed a self-administered questionnaire, and most were clinically examined for signs of hand eczema by the primary investigator at that time (A. Bregnhøj). previously validated questions were available. All questions were tailored to specifically focus on the past 4 years. Quality of life was measured by use of the Dermatology Life Quality Index (DLQI) (approved Danish translation) (15). The questionnaires were written in Danish, and, before approval, were tested among 10 Danish adult healthcare workers at Gentofte Hospital and among 14 enrolled year 4 hairdressing apprentices. Exclusion Participants were excluded if they had changed between intervention and control schools during their training, or if they had not answered the question pertaining to this subject. Definitions of outcome variables Hand eczema was defined according to an affirmative answer to the questions have you had hand eczema within the last 4 years? and/or have you had eczema on your wrists or forearms within the last 4 years? Severity of hand eczema was based on self-graded severity on a visual analogue scale (VAS) from 1 to 10. Employment status was based on the number of weekly working hours, and employment status was divided into three categories full-time employment ( 37 h/week), part-time employment (<37 h/week), or left the trade based on participants own statements. Evaluation of work habits for skin protection was based on how closely participants followed the guidelines given during their education regarding glove use in hours per day, for which functions gloves were used, whether gloves were reused inside out, hours of wet work per day, and whether participants cut hair before colouring it. Study population 6-year follow-up, 2015 The 284 apprentices who participated in the 18-month follow-up in the primary investigation were identified in the original dataset, and were sent a postal questionnaire in January 2015. Reminders with a new copy of the questionnaire were sent after 4 and 12 weeks. The questionnaire The questionnaire was composed of validated questions from the first and second follow-up during the primary investigation, and questions from the Nordic Occupational Skin Questionnaire (NOSQ-2002), to measure hand eczema prevalence and severity, and relevant exposure (14). New questions were formulated only when no Statistical analysis Statistical analysis was performed with SPSS for Windows (release V.18.0). The χ 2 -test was performed on cross tables with expected counts >5. Fisher s exact test was used on 2 2 tables when expected counts were <5. A test for trend was used for questions with ordinal categorized data. Student s t-test was used for comparison of group means. A p-value of <0.05 was considered to be significant. The incidence rate was calculated as the number of new cases divided by the population at risk during the last 6 years (participants reporting no hand eczema or incident hand eczema, n = 144); incidence rate = incident cases/(144 6years). Data protection agency permission: GEH-2014-046, I-Suite: 03240. Contact Dermatitis, 75, 32 40 33

Drop-outs n = 113 Drop-outs n = 127 Baseline n = 502 Intervention n = 301 Control n = 201 18-month Follow-up n = 284 Intervention n = 148 Control n = 136 6-year Follow-up n = 144 Intervention n = 72 Control n = 72 Excluded n = 105 Excluded n = 13 Fig. 1. Flow chart. Participants in the study from the intervention group and the control group. Results The questionnaire was sent to 284 participants from the primary investigation (7), and 157 responded (response rate 55.3%); 13 participants had either switched between intervention and control schools or did not answer the relevant question, and were therefore excluded. Accordingly, 144 participants were included in the current study: 72 from the intervention group and 72 from the control group (Fig. 1). Comparison of intervention group and control group The current participants answers from the primary investigation at inclusion, the 6-month follow-up and the 18-month follow-up were analysed (Table 1). No difference was seen in the prevalence of hand eczema at the 18- or the 6-month follow-up between the intervention group and the control group (p = 0.84 and p = 0.16, respectively). The intervention group included in the 6-year follow-up study had a significantly higher incidence of asthma and atopic dermatitis at the time of inclusion in the primary study than the control group. However, no difference was found between the groups in prevalence of hand eczema after participants with atopic dermatitis at inclusion had been excluded (results not shown). Comparisons of the control group and the intervention group at the 6-year follow-up showed no significant differences in the prevalence of hand eczema (Table 1), the severity of hand eczema (Table 2), and work routines (Table 4). Quality of life (Table 5) showed no difference between the groups in VAS score, but a significant trend was found when DLQI score categories Table 1. The prevalence of self-reported hand eczema, eczema on the hands, wrists, or forearms, and diagnosed atopic dermatitis Intervention group, % (n/n total ) Control group, % (n/n total ) p-value Total, % (n/n total ) Hand eczema Lifetime prevalence Eczema inclusion 9.7 (7/72) 5.6 (4/72) 0.35 7.6 (11/144) During education Eczema 6m 12.1 (8/66) 21.4 (14/66) 0.16 16.7 (22/132) Eczema 18m 22.2 (16/72) 23.6 (17/72) 0.84 22.9 (33/144) The past 4 years Eczema 6y 31.9 (23/72) 29.6 (21/71) 0.76 30.8 (44/143) New eczema cases 6y 47.8 (11/23) 47.6 (10/21) 0.99 47.7 (21/44) one-year prevalence 6y 23.6 (17/72) 21.1 (15/72) 0.72 22.4 (32/143) Point prevalence 6y 12.5 (9/72) 8.5 (6/71) 0.43 10.5 (15/143) Incident rate 18m to 6y Years per 1000 person-years 32.7 30.3 31.5 Hand eczema and/or wrist or forearm eczema Since last follow-up Eczema 6y 36.1 (26/72) 33.8 (24/71) 0.77 35.0 (50/143) 1-year prevalence 6y 27.8 (20/72) 22.5 (18/71) 0.47 25.2 (36/143) Point prevalence 6y 13.9 (10/72) 11.3 (8/71) 0.64 12.6 (18/143) Atopic dermatitis Baseline (lifetime prevalence) 23.6 (17/72) 9.7 (7/72) 0.025 16.7 (24/144) 6y (within the past 4 years) 11.4 (8/70) 4.3 (3/70) 0.12 7.9 (11/140) Comparison of the intervention group and the control group. Data are based on questionnaires from inclusion (inclusion), the 6-month follow-up (6m), the 18-month follow-up (18m), and the 6-year follow-up (6y), after exclusion as detailed in Materials and Methods. Self-reported eczema was validated in this cohort (7). The χ 2 -test was used for comparison of frequencies in the two independent groups. Fisher s test was used for expected counts < 5. Results with p < 0.05 were considered to be significant, and are in bold. 34 Contact Dermatitis, 75, 32 40

Table 2. Eczema severity and treatment Intervention group, % (n/n total ) Control group, % (n/n total ) p-value Total, % (n/n total ) VAS score at worst, 6y 5.85 (SD 2.92) 6.28 (SD 2.86) 0.60 6.08 (SD 2.87) Visits to dermatologist 44.0 (11/25) 26.1 (6/23) 0.20 35.4 (17/48) Patch tested 40.0 (10/25) 25.0 (6/24) 0.26 32.7 (16/49) Positive patch test reaction 90.0 (9/10) 42.9 (3/6) 0.06 75.0 (12/16) Topical steroid treatment 44.0 (11/25) 52.2 (12/23) 0.57 47.9 (23/48) Sick leave 16.0 (4/25) 8.7 (2/23) 0.38 12.5 (6/48) Caused or worsened by hairdressing work 80 (20/25) 91.3 (21/23) 0.24 85.4 (41/48) Reported as occupational disease 48.0 (12/25) 26.1 (6/23) 0.12 37.5 (18/48) SD, standard deviation; VAS, visual analogue scale. Comparison of participants with self-reported eczema on the hands, wrists or forearms in the intervention group and in the control group. Fifty participants experienced eczema on the hands, wrists or forearms during the past 4 years; the variation in total numbers is attributable to blank answers to some questions. The mean value of the VAS score is presented with the SD. A two-sample t-test was used to compare means of the VAS score between the groups. Table 3. Employment status, work hours, reasons for leaving the trade, and alternative occupations Intervention group, % (n/n total ) Control group, % (n/n total ) p-value Total, % (n/n total ) Employment Left the trade 39.4 (28/71) 29.6 (21/71) 0.22 34.5 (49/142) Hairdresser (all hours) 60.6 (43/71) 70.4 (50/71) 0.22 65.5 (93/142) Full-time ( 37 h/week) 65.1 (28/43) 84.0 (42/50) 0.035 75.3 (70/93) Employment after completing education Never worked/completed hairdressing education 10.0 (7/70) 11.3 (8/71) 0.81 10.6 (15/141) Considering leaving the trade 48.8 (21/43) 36.0 (18/50) 0.21 41.9 (39/93) Reasons for discontinuing Unfavourable working hours 32.1 (9/28) 45.0 (9/20) 0.36 37.5 (18/48) Back and joint pain 46.4 (13/28) 35.0 (7/20) 0.43 41.7 (20/48) Eczema on the hands, wrists, or forearms 17.9 (5/28) 5.0 (1/20) 0.19 12.5 (6/48) Occupations of those who left the trade Student 17.9 (5/28) 42.9 (9/21) 28.6 (14/49) Unemployed 14.3 (4/28) 23.8 (5/21) 18.4 (9/49) Childcarer/kindergarten teacher 14.3 (4/28) 14.3 (3/21) 14.3 (7/49) Sales assistant/salesperson 14.3 (4/28) 9.5 (2/21) 12.2 (6/49) Other professions 32.1 (9/28) 9.5 (2/21) 22.4 (11/49) Comparison of the intervention group and the control group. Full-time and part-time work were calculated from the question how many hours per week in the salon do you work on average. The χ 2 -test was used for comparison. Results with p < 0.05 were considered to be significant, and are in bold. Although 49 participants no longer worked as hairdressers, only 48 stated their reasons for leaving the trade. were compared. The control group s quality of life was more affected by their eczema than was the intervention group s. More participants in the control group (84.0%) than in the intervention group (65.1%) worked as hairdressers full-time, although the difference was not seen when participants with atopic dermatitis at inclusion were excluded. No difference was seen between the groups regarding leaving the trade (Table 3). At the 6-year follow-up, the intervention group still used gloves as often as seen at the 18-month follow-up. In contrast, in the control group, more participants used gloves at the 6-year follow-up for all work routines, although not all reached statistical significance (Fig. 2). Significant improvements in glove use were seen for rinsing out hair dye, using bleach, and using wave solution (p = 0.027, p = 0.036, and p = 0.006, respectively). At the 18-month follow-up, significantly more participants in the intervention group than in the control group (p = 0.039 and p = 0.017, respectively) used gloves for shampooing and rinsing out hair dye, a difference that was not found at the 6-year follow-up. The participants at the 6-year follow-up were considered as one group and described together regarding the following. Hand eczema. Of the participants, 50 (35%) had experienced eczema on the hands, wrists or forearms within the past 4 years; 44 (30.8%) had experienced hand eczema within the past 4 years, and, of these, 21 (47.7%) had Contact Dermatitis, 75, 32 40 35

% Golve users Work routines 100 90 80 70 60 50 40 30 20 10 0 Rinse Shampoo Rinse hair Rinse Wave Hair dye Bleaching wave ing dye bleaching solution solution Clean Eyebrow tools dye int 18m 42.9 98.4 87.3 87.3 76.2 49.2 76.2 12.7 1.6 con 18m 25.4 98.4 69.8 77.8 68.3 41.3 65.1 14.3 0 int 6y 40.7 98.3 91.5 91.5 86.4 49.2 74.6 10.2 1.7 con 6y 32.2 100 86.4 91.5 79.7 66.1 78 20.3 3.4 Fig. 2. Glove use among the intervention (int) group (blue) and the control (cont) group (red) at 18-month follow-up (dotted) and 6-year follow-up (full). The data include all participants from the 6-year follow-up study who worked as hairdressers after graduation. Significantly more participants used gloves for shampooing and to rinse-out hair dye in the intervention group than in the control group at the 18-month follow-up (p = 0.039 and p = 0.017, respectively). Significantly more used gloves for rinsing out hair dye, bleaching and wave solution in the control group at the 6-year follow-up than in the control group at the 18-month follow-up (p = 0.027, p = 0.036 and p = 0.006, respectively). There were no statistically significant differences between glove use in the intervention group at the 6-year follow-up and that in the intervention group at the 18-month follow-up. not had hand eczema in the initial study at the 18-month follow-up, which results in an incidence rate of 31.5 years per 1000 person-years. The 1-year prevalence of hand eczema alone was 22.4%; when eczema on the wrists and forearms was included, it was 25.2%. The point prevalence of hand eczema was 10.5%; when eczema on the wrists and forearms was included, it was 12.6% (Table 1). The eczema severity was scored on a VAS scale to an average of 6.08 [standard deviation (SD) 2.87] at its worst (Table 2). Of the participants with eczema on hands, wrists or forearms, 35.4% had consulted a dermatologist, and 12.5% had taken sick leave because of their eczema. When those with eczema on the hands, wrists or forearms were asked whether they believed that hairdressing affected their eczema, 41 of the 48 who answered the question (85.4%) stated that hairdressing was either the cause of their eczema or the reason for it worsening. Of the 48 who answered the question, 18 (37.5%) had reported their eczema to the National Board of Industrial Injuries (Denmark). The most common reasons for not reporting it were I thought it would go away (36.7%) and I wouldn t gain anything from it (26.7%). Employment status. Of the participants, 34.5% stated they had left the trade, and 10.6% had not completed the training to become a hairdresser (Table 3). Of those who still worked as hairdressers, 41.9% considered leaving the trade. Those who had left the trade stated that back and joint pain (41.7%) and unfavourable working hours (37.5%) were the main reasons for leaving; 12.5% reported eczema on the hands, wrists or forearms as a reason for leaving. Other reasons stated by the participants included loss of interest, low salary, airway allergies, and headaches. Of the 49 participants who had left the trade, 9 (18.4%) were currently unemployed and 14 (28.6%) were studying. The participants had changed to a wide variety of occupations. The two most dominant occupations were caregiver/kindergarten teacher (n = 7, 14.3%) and sale assistant/salesperson (n = 6, 12.2%). Other occupations included beautician, prison guard, zoo keeper, and jeweller. Allergy. Of the 144 participants, 21 (14.5%) stated that they had been patch tested within the past 4 years. Of those participants, 5 (23.8%) had positive reactions to nickel, 7 (33.3%) to perfume, and 4 (19.0%) to hair dye. 36 Contact Dermatitis, 75, 32 40

Table 4. The work routines among the apprentices who worked as hairdressers after graduation Intervention group, % (n/n total ) Control group, %(n/n total ) p-value Total, %(n/n total ) Glove use No gloves 6.3 (4/63) 6.3 (4/63) > 0.99 6.3 (8/126) Work routines with gloves Shampooing 40.7 (24/59) 32.2 (19/59) 0.34 36.4 (43/118) Hair dyeing 98.3 (58/59) 100 (59/59) 0.32 99.2 (117/118) Rinsing out hair dye 91.5 (54/59) 86.4 (51/59) 0.38 89.0 (105/118) Eyebrow dyeing 1.7 (1/59) 3.4 (2/59) 0.56 2.5 (3/118) Bleaching 91.5 (54/59) 91.5 (54/59) > 0.99 91.5 (108/118) Rinsing out bleaching solution 86.4 (51/59) 79.7 (47/59) 0.33 83.1 (98/118) Wave solution 49.2 (29/59) 66.1 (39/59) 0.062 57.6 (68/118) Rinsing out wave solution 74.6 (44/59) 78.0 (46/59) 0.67 76.3 (90/118) Cleaning tools 10.2 (6/59) 20.3 (12/59) 0.12 15.3 (18/118) Risk-associated work routines Turning gloves inside out 8.6 (5/58) 5.2 (3/58) 0.36 6.9 (8/116) Wet work (wet hands for 2 h) 66.7 (42/63) 71.4 (45/63) 0.56 69.0 (87/126) Wearing rings at work 57.1 (36/63) 62.9 (39/62) 0.51 60.0 (75/125) Preventive work routines Daily moisturizer 57.1 (36/63) 59.7 (37/62) 0.77 58.4 (73/125) Cut before colour 16.1 (10/62) 26.7 (16/60) 0.16 21.3 (26/122) Comparison of the intervention group and the control group. Turning gloves inside out includes hairdressers who state that they always, mostly or sometimes turn their gloves inside out at work. Wet work was stated as hours during an average work day. Wearing rings at work includes wearing finger rings always, mostly or sometimes when working. Daily moisturizer includes the use of moisturizer once daily or more. Cut before colour includes hairdressers who state that they usually cut a customer s hair before colouring it. The χ 2 -test was used for comparison. Work routines with gloves indicate percentage of hairdressers who always use gloves for the given procedure. When all participants were asked about reactions to allergen sources within the past 4 years, 39 (27.3%) stated that they had reacted to metal buttons and jewellery, and 19 (13.3%) had experienced eczema in the armpits after using deodorant containing perfume; 24.5% stated that they had experienced redness, scaling and itching after dyeing their own hair, and 6.9% reported reactions when dyeing their own eyebrows and/or eyelashes. Of participants reporting reactions to hair dye, 6 (4.2%) had experienced severe reactions, characterized as oedema of the eyelids and face, or wounds on the scalp. One of the 6 participants had been admitted to hospital as a result of such a reaction. Work routines. Work routines were analysed among the participants who had worked as hairdressers after completing the training (Table 4). Of the participants, 8 (6.3%) stated that they never used gloves when working; the majority of the remaining 118 participants stated that they always used gloves when applying hair dye (99.2%), when rinsing out hair dye (89.0%), and when applying and rinsing out bleach products (91.5% and 83.5%, respectively). Slightly fewer participants used gloves for applying and rinsing out wave solution (57.6% and 76.3%, respectively) than when handling hair dye and bleach products. When shampooing, 36.4% used gloves, but only 15.3% used gloves when cleaning tools in the salon; 8 (6.9%) stated that they turned gloves inside out to reuse them, and 69.0% stated that they had wet hands for >2 h on an average working day. Notably, 21.3% of all hairdressers stated that they mainly cut hair before colouring it, and 58.4% used hand moisturizer daily. Quality of life. The impact of hand eczema, including eczema on the wrists and forearms, on quality of life was measured according to the DLQI standard, and this showed that 41.7% of participants experienced skin problems that had a moderate to large effect on their quality of life, and none experienced an extreme effect (Table 5). When asked to grade their quality of life on a VAS scale, 10 being the best possible quality of life, the participants scoredameanof7.5(sd2.1). Discussion The current study was based on a previous controlled intervention study with a focus on the primary prevention of occupational hand eczema among hairdressing apprentices. In this 6-year follow-up study with 144 Contact Dermatitis, 75, 32 40 37

Table 5. The impact of skin problems on quality of life among participants with eczema on the hands, wrists, or forearms Intervention group, % (n/n total ) Control group, %(n/n total ) p-value Total, % (n/n total ) DLQI score DLQI 0 1 (no effect) 40.0 (10/25) 13.0 (3/23) 0.048 27.1 (13/48) DLQI 2 5 (small effect) 28.0 (7/25) 34.8 (8/23) 31.3 (15/48) DLQI 6 30 (moderate to large effect) 32.0 (8/25) 52.2 (12/23) 41.7 (20/48) VAS scale, quality of life Mean (1 = worst, 10 = best) 7.3 (SD 2.5) 7.7 (SD 1.8) 0.54 7.5 (SD 2.1) DLQI, Dermatology Life Quality Index; SD, standard deviation; VAS, visual analogue scale. Comparison of the intervention group and the control group. The DLQI score was calculated as described in the guidelines from Department of Dermatology Cardiff University School of Medicine. The χ 2 -test for trend was used for comparison of the index score (no effect, small effect, and moderate to severe effect). A two-sample t-test was used to compare means of the VAS score between the groups; mean values are shown with SD. Results with p < 0.05 were considered to be significant, and are in bold. The index measures the impact of all skin problems, including, but not limited to, hand eczema. former hairdressing apprentices, we were unable to detect a difference in the prevalence of hand eczema between the intervention group and the control group, both when examining the results from this cohort at the 6-year follow-up and when examining them at the 18-month follow-up. As a consequence of the relatively small number of participants in the current follow-up, the groups were no longer well matched, because they differed significantly in the prevalence of atopic dermatitis at inclusion. Atopic dermatitis has been described as a strong risk factor for developing hand eczema (11). This means that the intervention group included more susceptible persons. Nevertheless, hand eczema was not more prevalent than in the control group. This could be interpreted as a positive effect of the intervention, because even hairdressers with a history of atopic dermatitis were able to remain in the profession. The initial study was performed during the first 18 months of the 4-year hairdressing education. At that time, participants had a mean age of 17.8 years. At the 6-year follow-up, the participants had worked for <3 years as trained hairdressers, and were still considered to be a young population. Younger hairdressers have a higher prevalence of dermatological symptoms than their older colleagues (16), as hairdressers with skin problems often change career. The 1-year prevalence of hand eczema was 22.2% among all participants, which corresponds to a previous study among hairdressers in Denmark showing, in 2009, a 1-year prevalence of 20.7% among trained hairdressers (8). In comparison, matched controls in Denmark had a 1-year prevalence of 8.7% at the time of inclusion in the initial study (17). In the initial study at inclusion, 13.7% of all hairdressing apprentices had experienced eczematous reactions to hair dye, and, among a group of matched controls, 10% had reacted to hair dye (18). In comparison, there was a higher proportion of reported reactions to hair dye at the 6-year follow-up, as 24.5% reported reactions to hair dye in the past 4 years. A similar increase was not seen with reactions to jewellery, buttons, and other common nickel sources. The nature of the reported reactions is unknown, as only 21 participants reported being patch tested in the past 4 years. Of the patch tested individuals, 4 (19%) stated that they had been diagnosed with hair dye allergy. This shows that not only irritant contact dermatitis, but also allergies may appear relatively early in a hairdresser s career. Previous studies have found that hairdressers stay for only 8.4 years in the trade from starting their education (8). In our study, 35.5% had left the trade after 6 7 years, which agrees with the previous results. Fewer participants in the intervention group worked as hairdressers full time, which could be a sign of improved awareness of allergies and eczema, either because of the intervention or as a result of their experience with atopic dermatitis, as they also showed a tendency to more often state eczema as a reason for leaving the trade. This theory is further supported by the tendency for more of the intervention group to visit a dermatologist and more to report their eczema as an occupational disease, despite this not reaching statistical significance. Nevertheless, from this study, it is impossible to state whether the intervention group s increased awareness results from a life with atopic dermatitis or from the intervention itself; we suspect a combination of the two. When looking at exposure to irritants and allergens, we saw an improvement from the 18-month follow-up to the current study. Whereas the intervention group had glove use habits at the 6-year follow-up similar to those described at the 18-month follow-up (7), the control group had improved glove use for most work routines. The total percentages of participants using gloves were 38 Contact Dermatitis, 75, 32 40

increased for shampooing (36.4%) and rinsing out hair dye (89.0%) as compared with a 2009 study among Danish hairdressers in Denmark by Lysdal et al., in which only 10% used gloves for shampooing and 58% for rinsing out hair dye (19). Lysdal et al. also showed that 8% of the trained hairdressers reused their gloves by turning them inside out, which is similar to the 6.9% in our study. Another route of exposure to allergens and irritants from hair colour is cutting newly coloured hair without gloves (20); we were therefore pleased to find that 21.3% cut hair before colouring it. The limited number of participants makes it impossible to clearly identify small differences between the two groups, and introduces a possible selection bias, because those participating might over-represent either a healthy or a diseased part of the two groups. The study may also suffer from recall bias, as we relied on data from questionnaires. Most questions were directed at the past 4 years, and we hope to have limited the recall bias by using previously validated questions. Self-reported hand eczema has previously been validated in this cohort. It had a specificity of 99.8% and a sensitivity of 70.3% (21). The education part of the intervention was implemented in all hairdressing schools in 2011, the year after the 18-month follow-up. In the period between the 18-month and the 6-year follow-ups, there may well have been other initiatives to prevent hand eczema in the general population or at local workplaces. Taken together, these factors contribute to a general awareness of skin protection, resulting in our control group having access to the same information as that given by the intervention. Any difference between the groups is likely to be evened out over time as the information spreads to the control group. Conclusion In this 6-year follow-up, the groups (intervention and control) were no longer well matched. Many more participants with atopic dermatitis at baseline (23.6%) were present in the intervention group than in the control group (9.7%). Nevertheless, no increased prevalence of hand eczema was seen in the intervention group. In the intervention group, more participants worked part time than in the control group, but no difference was seen regarding leaving the trade. This may be an effect of the educational programme, as even individuals with a history of atopic dermatitis were able to remain in the profession. An overall improvement in work habits was also seen at the 6-year follow-up as compared with those of the participants in the initial study. We believe that a continued focus on good work routines could help to create a general shift in the attitude towards the hairdressers work environment, whereby protective routines are not seen as a weakness but something to take pride in. The goal remains to limit allergen and irritant exposure together with increased awareness of the initial signs of eczema and its optimal treatment. This could be achieved through further education of not only hairdressing apprentices, but also salon owners and other employers. Acknowledgements The study was financially supported by the Danish Hairdressers and Beauticians Union, the Danish Hairdresser Association, the Danish Working Environment Research fund, and the Aage Bangs foundation, which are all gratefully acknowledged. References 1 Lind M L, Albin M, Brisman J et al. Incidence of hand eczema in female Swedish hairdressers. Occup Environ Med 2007: 64: 191 195. 2 Cahill J W J, Matheson M, Palmer A et al. Occupational contact dermatitis: a review of 18 years of date from an occupational dermatology clinic. Report for Safe Work Australia Australia, 2012. 3 Behroozy A, Keegel T G. Wet-work exposure: a main risk factor for occupational hand dermatitis. Saf Health Work 2014: 5: 175 180. 4 Søsted H, Rustemeyer T, Gonçalo M et al. Contact allergy to common ingredients in hair dyes.contact Dermatitis 2013: 69: 32 39. 5 Hansen H S, Søsted H. Hand eczema in Copenhagen hairdressers prevalence and under-reporting to occupational registers.contact Dermatitis 2009: 61: 361 363. 6 Uter W, Gefeller O, John S M et al. Contact allergy to ingredients of hair cosmetics a comparison of female hairdressers and clients based on IVDK 2007 2012 data. Contact Dermatitis 2014: 71: 13 20. 7 Bregnhøj A, Menné T, Johansen J D, Søsted H. Prevention of hand eczema among Danish hairdressing apprentices: an intervention study. Occup Environ Med 2012: 69: 310 316. 8 Lysdal S H, Søsted H, Andersen K E, Johansen J D. Hand eczema in hairdressers: a Danish register-based study of the prevalence of hand eczema and its career consequences. Contact Dermatitis 2011: 65: 151 158. 9 Hougaard M G, Winther L, Søsted H et al. Occupational skin diseases in hairdressing apprentices has anything changed? Contact Dermatitis 2015: 72: 40 46. 10 Roberts H, Frowen K, Sim M, Nixon R. Prevalence of atopy in a population of hairdressing students and practising hairdressers in Melbourne, Australia. Australas J Dermatol 2006: 47: 172 177. 11 Uter W, Pfahlberg A, Gefeller O, Schwanitz H J. Hand dermatitis in a prospectively-followed cohort of hairdressing apprentices: final results of the POSH study. Prevention of Contact Dermatitis, 75, 32 40 39

occupational skin disease in hairdressers. Contact Dermatitis 1999: 41: 280 286. 12 Held E, Mygind K, Wolff C et al. Prevention of work related skin problems: an intervention study in wet work employees. Occup Environ Med 2002: 59: 556 561. 13 Clemmensen K K, Randboll I, Ryborg M F et al. Evidence-based training as primary prevention of hand eczema in a population of hospital cleaning workers. Contact Dermatitis 2015: 72: 47 54. 14 Susitaival P, Flyvholm M A, Meding B et al. Nordic occupational skin questionnaire (NOSQ-2002): a new tool for surveying occupational skin diseases and exposure. Contact Dermatitis 2003: 49: 70 76. 15 Finlay A Y, Khan G K. Dermatology Life Quality Index (DLQI) a simple practical measure for routine clinical use. Clin Exp Dermatol 1994: 19: 210 216. 16 Jung P K, Lee J H, Baek J H et al. The effect of work characteristics on dermatologic symptoms in hairdressers. Ann Occup Environ Med 2014: 26: 13. 17 Bregnhøj A, Søsted H, Menné T, Johansen J D. Healthy worker effect in hairdressing apprentices. Contact Dermatitis 2011: 64: 80 84. 18 Bregnhøj A, Sosted H, Menné T, Johansen J D. Exposures and reactions to allergens among hairdressing apprentices and matched controls. Contact Dermatitis 2011: 64: 85 89. 19 Lysdal S H, Johansen J D, Flyvholm M A, Søsted H. A quantification of occupational skin exposures and the use of protective gloves among hairdressers in Denmark. Contact Dermatitis 2012: 66: 323 334. 20 Lind M L, Boman A, Sollenberg J et al. Occupational dermal exposure to permanent hair dyes among hairdressers. Ann Occup Hyg 2005: 49: 473 480. 21 Bregnhøj A, Søsted H, Menné T, Johansen J D. Validation of self-reporting of hand eczema among Danish hairdressing apprentices. Contact Dermatitis 2011: 65: 146 150. 40 Contact Dermatitis, 75, 32 40