ASSESSMENT OF FACIAL WRINKLES 133 Table III Consensus Values of Evaluation by Five General Examiners of Severity of Wrinkles at the Outer Eye Corners of 247 Subjects Wrinkle grade n (%) 1 36 (14.6) 2 25 (10.1) 3 21 (8.5) 4 36 (14.6) 5 33 (13.4) 6 35 (14.2) 7 25 (10.1) 8 19 (7.7) 9 17 (6.9) Total 247 Table IV Inter-Observer Agreement on Eye Corner Wrinkles Using the Five- and Nine-Grade Photoscales 5-Grade photoscale 9-Grade photoscale Examiner fo/ N.r* K fo/NT* YT 86/164 0.407 130/247 0.464 TF 89/164 0.426 114/247 0.397 KT 114/164 0.604 139/247 0.507 AO 84/164 0.400 108/247 0.361 HO 122/164 0.657 83/247 0.252 * Number of scores in agreement with the consensus panel (fo) divided by total number (NT). Table V Reproducibility of the Nine-Grade Photoscale Scores of Eye-Corner Wrinkles of 247 Subjects Exp. 1 Exp. 2 Examiner fo/N•r* K fo/NT* • YT 130/247 0.464 122/247 0.425 TF 114/247 0.397 113/247 0.392 KT 139/247 0.507 127/247 0.452 AO 108/247 0.361 132/247 0.474 HO 83/247 0.252 99/247 0.327 * Number of scores in agreement with the consensus panel (fo) divided by total number (NT). AGE-RELATED CHANGES IN THE WRINKLE SCORES AT EACH FACIAL SITE The age-related changes in the five-grade scores at each facial site are shown in Table VIII. The correlation coefficient and the slope of the linear approximate expression were calculated. The sites showing a high correlation with age were as follows, in the order shown: eye corners lower eyelids upper eyelids cheeks forehead mouth angles nasolabial grooves glabella. The relationships between age and the five grades at representative sites (eye corners and cheeks) are shown in Figure 3 and Figure 4, respectively. The slope of the linear approximate equation at each site is shown in Figure

134 JOURNAL OF COSMETIC SCIENCE Table VI Distribution of Forehead Wrinkle Scores According to the Five-Grade Scale Wrinkle grade n (%) 1 54 (32.9) 2 24 (14.6) 3 35 (21.3) 4 34 (20.7) 5 17 (10.4) Total 164 Table VII Consistency of Agreement Regarding the Eye-Corner Wrinkles and Forehead Wrinkles of 164 Subjects, Scored by Five General Examiners Eye corners Forehead Examiner fo / NT* K fo / NT* K YT 86/164 0.407 79/164 0.328 TF 89/164 0.426 79/164 0.322 KT 114/164 0.604 93/164 0.441 AO 84/164 0.400 87 / 164 0.391 HO 122/164 0.657 79/164 0.332 * Number of scores in agreement with the consensus panel (fo) divided by total number (NT). 5, and the ages at which the wrinkle score obtained from this slope became 3 are shown in Figure 6. The slope was high, in the order of: eye corners forehead lower eyelids upper eyelids mouth angles nasolabial grooves cheeks glabella. The ages at which the score became 3 were the early 30s for the eye corners and lower eyelids, the late 40s for the upper eyelids, nasolabial grooves, and forehead, the late 40s for the glabella and the mouth angles, and the middle 50s for the cheeks. CLINICAL TEST OF WRINKLE-IMPROVING AGENT The demographic details of the subjects in the clinical test are shown in Table IX. Their mean age was 42.4 years, and scores 6 on the nine-grade scale were observed in 33%. The results of the clinical test are shown in Table X. The wrinkle scores at the eye corners were significantly improved after the use of the agent for eight weeks. DISCUSSION In a preliminary study, the results using the photo scales were compared with those using descriptive standards, and the usefulness of the photo scales was confirmed (data not shown). Griffiths et al. (8) evaluated photodamage according to photo scales and descriptive standards and found the kappa value significantly higher for the former (0.31) compared to the latter (0.11), suggesting the usefulness of photo scales. The results of our preliminary evaluation were consistent with their findings. Here, a five- grade photo scale was compared with a nine-grade scale, and the former scale was