20 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS area reaches a certain level (around 1000 Ixm2), it no longer appears to limit percuta- neous absorption and TEWL. On the contrary, when corneocyte surface area decreases, TEWL and percutaneous absorption increase, but only to a limit. Finally, except in the elderly (point no. 3), inward movement of a non-water-soluble compound such as ben- zoic acid and outward movement of water have the same type relationship to corneocyte size. Irrespective of anatomic site and sex, a linear relationship (r = 0.92, p 0.001) exists between total penetration of benzoic acid and TEWL (Figure 4). (Point no. 3 corresponds to subjects aged 65-80 measured on the upper-outer arm. This case is special and will be discussed. It was not taken into account in the calculation of the correlation coefficient). DISCUSSION CORNEOCYTE SIZE-SKIN PERMEABILITY RELATIONSHIP Influence ofanatomic site. The forehead and the postauricular region exhibit greater per- meability than the arm, the forearm, or the abdomen. Since the thickness and number of cell layers of the stratum corneum of all the sites involved in this study are similar (on average 12 Ixm and 18 cell layers) (19,20), the differences in permeability that we have observed have to be explained by additional structural and physicochemical criteria. Thus it is conceivable that the higher penetration through areas such as the forehead might be explained by an increase in transfollicular in preference to transepidermal absorption. However, how can one explain the large difference (a factor of 50 to 100) that exists in the number of sebaceous glands between the arm and the forehead (21,22) 11oo lOOO 900 800 700 600 5OO ,i. 6 10 20 30 5 10 Total penetration of benzoic acid within 4 days (nmol. cm 'z) Transepidermal water loss ( T. (•.W.L.) (gm.m -z.h 'l) Figure 3. Relationship between (A) percutaneous absorption of benzoic acid and (B) transepidermal water loss (TEWL) and corneocyte surface area according to anatomic site, age, and sex in man. 1. Arm (upper- outer, male, 20-30 yr). 2. Arm (upper-outer, male, 45-55 yr). 3. Arm (upper-outer, male, 65-80 yr). 4. Arm (upper-outer, female, 20-30 yr). 5. Abdomen (male, 20-30 yr). 6. Postauricular (male, 20-30 yr). 7. Forehead (male, 20-30 yr). 8. Forehead (female, 20-30 yr). 9. Forearm (ventral-elbow, male, 20-30 yr). 10. Forearm (ventral-mid, male, 20-30 yr). 11. Forearm (ventral-wrist, male, 20-30 yr). (C) according to Marks et al. (11).

CORNEOCYTE SIZE AND PERMEABILITY 21 30 20 10 r = 0.92 e 0.001 6 l/T 7 5 10 Transepidermal water loss (TEWL) (g. m -2. h -• ) Figure 4. In vivo relationship between transepidermal water loss (TEWL) and percutaneous absorption of benzoic acid according to anatomic site, age, and sex, in man. 1. Arm (upper-outer, male, 20-30 yr). 2. Arm (upper-outer, male, 45- 55 yr). 3. Arm (upper-outer, male, 65-80 yr). 4. Arm (upper-outer, female, 20-30 yr). 5. Abdomen (male, 20-30 yr). 6. Postauricular (male, 20-30 yr). 7. Forehead (male, 20-30 yr). 8. Forehead (female, 20-30 yr). 9. Forearm (ventral-elbow, male, 20-30 yr). 10. Forearm (ventral- mid, male, 20-30 yr). 11. Forearm (ventral-wrist, male, 20-30 yr). and the relatively slight difference (a factor of 3) in the penetration of benzoic acid between these two areas? The permeability of the postauricular area towards both water and benzoic acid is close to that of the forehead. From the morphologic standpoint, the closeness of the capil- laries to the surface of the skin may promote the resorption of substances and give the postauricular area its high permeability (23). For common sites, the rank order in corneocyte size established by Piewig et al. (17) and Marks et al. (11) is confirmed in our work. Thus, in areas such as the abdomen, the upper-outer arm, the ventral-mid and ventral-elbow forearm and the abdomen, there is no significant difference in corneocyte size. It is now accepted that an inverse correlation between epidermal cell proliferation and corneocyte size exists (24). Moreover, it has been suggested that there might be differences in the epidermal turnover according to anatomical location (25-28). Note (Table I, Figure 2) that anatomic sites in which corneocyte size does not vary are all areas which are more or less protected by clothing. On the contrary, in regions such as the wrist, the postauricular area, or the forehead, more exposed to environmental factors (including sunlight), corneocytes are smaller. Taking into account the fact that in all the sites tested tb.e total thickness of the stratum corneum (19) and the individual thickness of the corneocytes (29) do not change, the relationship between the flat area of the horny cells and the barrier function of the stratum corneum must be addressed. In permeability phenomena, the current trend is to assign priority to intercellular rather