Prevalence of hand dermatoses among Finnish farmers.

P, K. Prevalence of hand dermatoses among Finnish farmers. Scand J Work Environ Health 1994;20:206-12. OBJECTIVES - The goal of this study was to determine the prevalence and risk factors of hand der matosis among farmers. METHODS - A questionnaire survey was carried out in a geographically defined sample of the Finn ish farming population between the ages of 18 and 64 years (N = 10847). RESULTS - The one-year prevalence of self-reported hand and forearm dermatoses was 16% for the women and 7% for the men. These figures were similar to the prevalence of hand eczema in two large Scandinavian questionnaire surveys. The highest one-year prevalence of hand dermatoses was found for women on farms with more than nine dairy cows (20%). Atopy (both the personal history of atop ic dermatitis and respiratory atopy), female gender, and, among the women, also age under 35 years were the most important risk factors for the occurrence of hand dermatosis. Work-related risk factors were handling disinfectants daily, handling silage preservatives, milking cows, and machine servic ing. CONCLUSIONS - Occupationalrisk factorsfor handdermatoses werefound,especiallyfor dairyfarming. The results may be useful for the prevention of hand dermatoses in farming since they direct atten tion to related occupational hazards, and they may also aid vocational guidance for the atopic popula tion. In order to lower the prevalence of hand dermatoses in dairy farming, both preventive and pro tective measures should become everyday practice in farming work.

The aims of this study were to determine the occurrence of hand dermatoses among Finnish farmers and to establish which risk factors are constitutional (gender, atopy) and which are work-related.

Subjects and methods
In November 1979 a questionnaire was sent to all of the 14 346 registered 18-to 64-year-old farmers in 14 Finnish municipalities. The sample was 5% of all farmers in Finland, and it represented the geographic, social, gender, and age distributions and the different farm operations of the total farming population (6). The questionnaire was self-administered and contained personal questions on health and work.
A total of 13 054 farmers filled out and returned the questionnaire (91 %). Of these, 10 847 were professional farmers, who constituted the population of this study. The mean ages in the study population were 44.7 years for the men (N = 5581) and 46.1 years for the women (N = 5266). Thirty percent of the men and 26% of the women were under 40 years of age. The study included 6796 farms. The mean area of cultivated land was 13.2 ha per farm. On most farms, a married couple (family) or a single farmer did all of the work without hired workers. On 4484 farms (66%) with 7508 farmers (69%) there was at least one dairy cow. The mean number of dairy cows was 7.5 per farm. On 6% of the farms the main farm operation was swi ne breeding, on 4% it was other animal breeding, mostly beef cattle, and on 14% it was raising crops , other plant cultivation or fore stry. On the farms with dairy cattle, 86.3% of thc women and 56 .2% of the men milked cow s regul arly. Milk ing was still done by hand on 51 % of the farms with one to five cows in 1979.
In the questionn aire, the first two questions concerning the skin were "Do you have a skin disea se now ?" and "Have you had any sk in disease within the last year?" The positive answers to the first que stion gave the point pre valence, and positive answers to either of the questions provided the oneyear prev alence of skin diseases . The persons who answered "yes" to either of the two questi ons were asked to shade the area s of their eczema on a picture (figure I). The hand and forearm were togeth er in the same area. Data on skin di sease areas were missing for 5.9 % of the men and 5.6 % of the women . The questions about atopic background were "Do you have or have you had infantile eczema or eczema on antecubital fossae or knee flexures?" and " Do you have or have you had seasonal or other aller gic rhinitis or bronchi al asthm a?" Farmer s with a positive answer to the first que stion formed the "atopic dermatitis" group , which also included people with respiratory symptoms. Farmers with a positive answer to the second questi on alone were included in the "atopic respiratory disea se" group . The two groups were includ ed in the atopies, and the rest of the subjects were con sidered nonatopi c. A history of atopic dermatitis was reported by 9.8 % of the men and 10.2% of the women, and the percentage of farmers with a history of allergic rhinit is or asthma was 22.3% for the men (18 .7% atopi c respiratory disease alone) and 23.9% for the women (19.9 % atopic respiratory disease alone ). Some atopic symptoms were reported by 29.3 % of the farmers.
Gender, age, and famil y size (all peopl e living in the household) were analyzed from the demographic factors. The covariate s from the work were selected with the following three criteria: (i) the type of work (dairy cattl e, swine breed ing, other animal breeding , plant cultivation), (ii) everyday jo bs known to be hazardous to skin (milking, handling disinfectants, handling silage preservatives, machin e service work), and (iii) the total work load (the area of cultivatcd land and the number of dairy cows). All subjects with at least one dairy cow were in the da iry farmer group.
In the statistical analysis of the data, cross -tabulation with the chi-square test, both for heterogeneity and trend, and the multivariate logisti c linear regre ssion model were used. Farm oper ation and number of dairy cows were used separately in the logistic regression model because they measured partly the same parameters (cattle). Farm ers with missing data on dermatoses were included in the group with no self-report ed dermatoses. The logis- tic regression and the calculation of risk ratios were performed by general linear interact ive modeling (7).

All dermatoses
The point prevalence (pre sent dermatosis) of all reported dermatoses was 16.2% for the men and 19.8% for the women. The one-year prevalence of all dermatoses was 23.4% for the men and 29.4 % for the women (figure 2). The skin disease had persisted for at least one year prior to the inqu iry for 42% of the men and 48% of the women with dermatoses.
In the atopic dermatiti s group, the one-y ear prevalence of all recent or present dermatoses was more than twice of that in the nonatopic group, and the prevalenc e was also higher in the atopic respira tory disease group than in the nonatopic group (P < 0.00 I) (figure 2). The number of atopies among the farmers with dermatoses was 41 % for the men (20.4% with atop ic dermatitis and 20.4 % with atopic respiratory disease) and 44% for the women ( 19.3% with atop ic derm atit is and 25. 1% with atopic respirat ory disease).
The hands and forearm s were the most frequent sites of reported dermatoses among both genders. The second highest prevalences were reported for the leg. The female farmer s reported dermato ses on their hands or forearms (P < O.OOl) and face (P< O.OOI) more often than the male farmer s, who reported der- matoses on their trunk (P<O.OI) and legs (P<O.OI) more often than the female farmers (figure I).

Hand and forearm dermatoses
The women reported hand dermatoses more often than the men (P < 0.00 I .) The number of dairy cows showed a positive correlation with the prevalence of hand dermatoses, as did regular milking also (table I). Half of the nonatopic men handled silage preservatives, and half of them did machine service work. Both types of work correlated positively with the prevalence of hand dermatoses (table I). 208 Female farmers. The prevalence of hand dermatoses decreased also among the female farmers after the age of 34 years (P<O.OOI, X 2 for trend). The female farmers working on farms with dairy cattle had hand dermatoses more often than the women on other farms (table 1). The number of dairy cows and the area of cultivation showed a positive correlation with the prevalence of hand dermatoses (table I). A positive correlation was also seen between hand dermatoses and regular milking (table I). The women with more than three family members had more hand dermatoses than those with smaller families (P < 0.00 I, X 2 for trend). Forty-nine percent of the nonatopic female farmers handled disinfectants daily, and they had more hand dermatoses than those handling disinfectants less often or not at all (table I). The daily handling of disinfectants correlated also with increasing numbers of cows and family members. Of the nonatopic women 35% handled silage preservatives; this activity also correlated positively with the prevalence of hand dermatoses in this group (table I).
Atopy groups Male farmers. Male farmers with atopic respiratory disease had hand dermatoses in swine breeding significantly more often than farmers in other operations (table 1). The number of dairy cows and the area of cultivation correlated also positively with the prevalence of hand dermatoses among the male farmers with atopic respiratory disease.
Female farmers. The prevalence of hand dermatoses decreased by age among female farmers with atopic respiratory disease (P < 0.05, X 2 for trend). Hand der- matoses showed a posrtive correlation with increasing number of dairy cows in the atopic dermatitis group. Fifty-one percent of the women with atopic dermatitis handled disinfectants daily, and 41% of them (42% in the atopic respiratory disease group) handled silage preservatives. Both types of activity increased the prevalence of hand dermatoses in the atopic dermatitis group, the latter also in the atopic respiratory disease group (table I).

Combined effect of constitutional and occupational risk factors on the occurrence of hand dermatoses
The effects of risk factors on the prevalence of hand dermatoses were calculated in a linear regression model as risk ratios (RR). The results are listed in table 2. The risk of hand dermatosis was about threefold for the farmers with atopic dermatitis, and for the farmers with atopic respiratory disease there was almost twice the risk of the nonatopic group. For the female farmers, age under 35 years was also a risk factor for hand dermatoses, Significant risk ratios were also found for the risk of hand dermatoses among the male farmers who handled silage preservatives and among the female farmers who handled disinfectant detergents daily (  Table 2. Risk rat ios (RR) and their 95% confidence intervals (95% Cl) for the effects of the covariates on the one-year prevalence of hand dermatoses among Finn ish farmers (558 1 men and 5266 women). Analyses were separate for the genders. The models included the listed covariates. Deviance of the models : 634 .0 (men) and 526.6 (women), deg rees of freedom: 754 (men) and 471 (women). On ly 38 women did machine servicing, and 13 of them reported hand derm atosis.

Discussion
The reported lifetime inc idence of atop ic dermatitis was 10%, that of respiratory atopy 23 %, and that of any atopic sy mpto ms 29 % in our study popu lation.
In other Sca ndi navian pop ulation studies (4,5,(8)(9)(10)(11)(12)(13), the lifetime incidences of atopic symptoms have been so mewhat lower than or similar to those of this study. Our fig ures do not suggest that atopic farmers woul d have left the farming industry. Our po int prevalences for all reported der matoses were com parable with the respective figures in four other Scan dinavian postal surveys (10, 12, 14 , 15). The poi nt preva lence for skin diseases varied in these surveys between 16 and 24 (mean 18)%. The point prevalence for sk in diseases ranged between 16 and 3 1 (mean 2 1)% in clinical popu latio n studies (8,(16)(17)(18). 210 The point prevalences of hand eczema have varied between 8 and 14% in most cli nica l studies (12 , 15, 19-22), but also very different figures (1.7-83%) have been found (23)(24)(25). There are neither uniform criteria nor stan dards for the term eczema or der matosis in clinical studies or in postal surveys; the result is a large variation in the prevalence figures. The po int prevalence of self-reported hand dermatoses amo ng the wome n of our study was somewhat hig her (11 %) than that of cl inical hand eczema among Lamrni nta usta 's hospital workers (8 %) ( 19). The one-year prevalence of self-reported hand eczema has varied fro m 5 to 9% among men and 11 to 15% among women in studies based on postal surveys of general populatio ns (26)(27)(28). The correspondi ng figures in the present study (7% for the men and 16% for the women) wer e simi lar to these prevalence figures.
The reliability and reproducibility of survey methods for epidemiolog ic studi es of skin diseases have not bee n estab lished in the literature. In add ition, there are no standardized questionnaires for dermatologic research. The similarity in atopy prevalences in the present material with those of earlier studies allows us to assume that a questionnaire survey can detect atopic symptoms fair ly relevantly. When the same questionn aire as in the present study was used three years later (in 1982) on the same population, similar prevalences of hand dermatoses were encountered, the poin t prevalence being 10.9% for the wome n and 4.9 % for the men and the one-year prevalence bei ng 16.8% for the women and 7.7% for the men (unpublished res ults) . This finding suggests good reproducibility for the present method. Meding & Swanbeck (27) found hand eczema in 89 % of the persons who had stated that they had it ("considered themse lves to have had hand eczema on some occasion during the pre vious 12 mo nths") in the questionnaire survey. Thi s finding sugges ts goo d reliability for the method, which was simi lar to ours. In the same Swe dish study (5) , the prevalences of hand dermatoses for persons with a history of no personal atopy and for persons with a history of ski n atopy were striki ngly similar to our fig ures. This fact also favors the concept that prevalences of dermatoses and atopy can be inves tigated with questionnaires.
About 50 % of the farmers reporting hand dermatoses in the questionnaire also reported atop y. This figu re is similar to that of the Swedish population study by Meding & Swa nbeck (5). The personal history of atop ic de rmatitis increased the risk for hand derm ato sis abo ut threefold in the present study (figure 2) . The prevalence of hand eczema has been shown to be increased among peo ple with a personal history of atop ic dermatitis in many studie s (4,5,19,(29)(30)(31).
There is some con troversy about the risk of hand eczema in respirat ory atopies without the history of atopic dermatitis. The history of respiratory atopy almost doubled the risk for hand dermatosis among both the men and the women in our study (table 2). Atopic respiratory disea se was the stronges t risk factor for hand dermatosis among the male farm ers without other sites of derm atosis. Th ese findings agre e with tho se reported by Nil sson (4), Med ing & Swanbeck (5) , and Kri sten sen (3 1). Rystedt (3) reported that, in his study, the prevalence of hand eczema among people with a personal history of asthma equaled th at of the nonatopic referents, being abo ut 5%. In Lammintausta & Kalimo's study (II ) of hospital person nel in wet work cumulative res piratory atopy alone without dry ski n (atopic ski n diathesis) did not increase the prevalen ce of hand ecze ma . In Rystedt' s study there was also a stro ng corre lation bet ween dry and itch y skin and hand eczema when there was occupatio nal exposure to moisture, det erg ents, or chemica ls (3) . The result s were si milar in a study by Fartasch et al (32), wh o found atopic skin diathesis but not re spiratory atopy to be a significant risk factor for hand eczema in thei r logistic regre ssion . These resul ts are in accordance with those of a study by Werner (33) , who found that only cli nically dry skin of pat ients with atopic dermatitis had a lower wa ter conte nt than referents. Ski n sensitiv ity (atopic skin diathesis) was not registered in the pre sent study or in other studies where the history of resp iratory atopy as such has been co nnected with higher pre valen ces of hand eczema (4 , 5, 3 1).
In Meding & Swanbeck ' s (5) questionnaire study of an urban population the prevalence of hand dermatoses in the nonatopic group and the atopi c dermatitis group were very simi lar to tho se of the present study, but there wer e s ignifica ntly fewer cases of hand der matoses in the atopic respiratory disease gro up of urban Swedes than in those of Finnish fa rmers ( 1104 versus 15.5%). Thi s find ing may be explained by the farm ers' occupational expos ure to respiratory allerge ns, which ca n also ca use skin pro blems (eg , ani mal dander, flo urs). Th e prevalence of all dermatoses in the present study was also significantly higher for the atopic respiratory disease group than for the nonatopic group, among both the men and the women (figure 2).
On Finnish farms in 1979, wome n did most of the milking and were exposed to wet and dirty work, disinfcc tants, and allerge ns (anima l dander, rubb er, etc) more than men . Among fe male farm er s, the risk for hand dermatoses was more than twofold, and for hand derm atoses without other sites of derm atosis the risk was fou rfold that of the men ' s. The risk for hand derm atiti s was higher for wo men und er 35 yea rs of age than for older women. There are also other studies that sho w higher hand ec ze ma prevalen ces for women than men, and also hig her prevalences for young women (20,26,27) . In addition to domestic wor k and other environmenta l ex posures , constit ution al factor s of wome n may be in vol ved .
Occup ational risk factors were found especially in dair y farming (ha ndling di sinfect ant s, handl ing si-Scand J Work Environ Health 1994, vol 20, no 3 lage, and milking) and machine serv icing, but they were weak compared with the con stitutional on es. Am ong the men with hand dermatosis, but witho ut oth er sites of derm atosis, milking was almost as stro ng a risk factor as atopy. When interaction s with atopy and work were analyzed, atop y did not increase the effect of the work-related risk factors on the prevalen ce of hand dermatoses. The illogical finding that men with atopic dermatitis who milked regularly had less hand dermatoses th an those who did not milk (table I ) was ob viou sly due to the self-ex clusion fro m milking wor k by those with hand derm ato ses.