SPECIFIC LANOLIN ALLERGY 327 products not present in the original lanolin. It is thus not certain that a positive reaction to lanolin alcohols proves sensitization to lanolin. INHERENT EXAGGERATION OF TEST RESULTS Even if both (a) and (b) modifications are employed in the same test series, since each has its own uncertainties the results must still be incon- clusive and much of the recently published work can only be regarded as presenting an exaggerated picture of the situation. Different opinions on the extent of exaggeration have been given, but according to Epstein (22) the addition of salicylic acid to lanolin or to eucerin can yield positive results of which half or more are false. Epstein also considers that patch testing with 30• of lanolin alcohols in petroleum jelly produces a significant number of false-positive irritant reactions, and would regard only positive reactions to 20• and 10• concentrations as 'clear cut indicators of lanolin allergy'. This is a helpful qualification but does not resolve fully the known uncertainties. Differences in dermatological effect between lanolin and lanolin alcohols have, in fact, been established by de Beukelaar (23) and by Bandmann and Reichenberger (24). There are a number of quantitative comparisons of the incidence of positive reactions from lanolin alcohols and lanolin respectively. Thus Reichenberger (25) found diluted lanolin alcohols to elicit fifteen times as many reactions as lanolin, whilst Wereide (9) found 7.5 times as many when both test substances contained 5• of salicylic acid. Thune (2) found 10 times as many with added salicylic acid, and 3.2 times without. An increase by a factor of 3.7 was reported by Hjorth and Trolle-Lassen (10), a factor of 2.5 by Baer, Serr and Weissenbach-Vial (31) and a factor of 5.5 by Epstein (22). When a keratolytic was added to lanolin, Thune (2) demonstrated increases in the incidence of positive reactions by factors of 2.5 and 2.6, and he attributed this exaggeration to false positive reactions whilst also believing that even 2• of added salicylic acid caused too many false positives and that salicylic acid in petroleum jelly used as a control could yield false negatives since the jelly is not as penetrating as lanolin. Again on adding a keratolytic, Hjorth and Trolle-Lassen (10) found a factorial increase in incidence of 2.8, and Epstein (22) a factor of 3.0. Fisher (8) emphasized that tests for lanolin sensitivity should be carried out with 100 per cent lanolin alone, since admixture with keratolytic, emulsifier or anything which increased penetration into the skin could cause false positive results.

328 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Although these weaknesses and criticisms of both test modifications are severe, it must be recognized that such powerful methods have been used by investigators because of the difficulty of diagnosing hypersensitivity to a weak allergen, and may well be a help, if not essential, in making the first critical diagnosis of a troublesome case. A utoxidation The pattern of allergic response to lanolin is irregular. It can vary between age groups and sexes (10), and between different localities (27), (28). Some samples of lanolin or lanolin alcohols have elicited positive reactions, whilst others have not. The possibility of this difference in be- haviour being due to the presence or absence of autoxidafion by-products has been investigated and significant evidence found to the effect that not only did autoxidation of lanolin alcohols markedly increase the incidence of allergy, but also that the addition of certain antioxidants to lanolin depressed the level of incidence (10). Schwarzreid (26) referred to the possibility of autoxidation degradation products being responsible for allergic reaction to lanolin, and for a similar reason Reichenberger (25) recommended testing with both fresh and aged eucerin. As a result of research into their effects (29), (30), the addition of ap- proved antioxidants to lanolin is now permitted by a number of pharma- copoeiae, an addition to Wool Alcohols being mandatory in the BP since 1968. For this reason one might expect the general incidence of allergic hypersensitivity to lanolin and lanolin alcohols, whatever this incidence is, to be favourably influenced. CROSS-SENSITIZATION AND MULTIPLE SENSITIVITY Efforts have been made to relate allergy to lanolin with concomitant allergy to other substances such as long chain fatty alcohols, isopropyl myristate, Lanette Wax and glyceryl monostearate (10), (12). No specific or constant relationship could be established, but many cases of multiple sensitivity were found. For example, Magnusson et al. (27) tested six groups of patients at different hospitals and listed the percentage within each group of positive reactions to a total of 24 different substances. The column totals varied from 11•o to 158•o, indicating a considerable degree of multiple sensitivity. In the work of Fregert et al. (28), the percentages total 78.5•o but this was the proportion of 4825 patients out of whom only 40• gave a