392 JOURNAL OF COSMETIC SCIENCE shift in dermal fluid from the face toward the legs by gravity over the course of each day, resulting in the elicitation of a diurnal variation in skin thickness (12). It is thus conceivable that the diurnal variation in skin thickness is also modulated by aging as a result of differing degrees of dermal fluid accumulation between the upper and lower body. To clarify the effects of diurnal variation on age-related changes in skin thickness, we have now studied skin thickness in the morning and in the afternoon by B-mode ultrasonography at three sites on the face (the forehead, the corners of the eye, and the cheek) as sun-exposed areas and on the ventral side of the forearm as a poorly sun-exposed area. Here we report that there is an age-dependent slight increase in the thickness of all the facial skin sites measured in the morning, which may be primarily ascribed to an increase in an age-related accumulation of dermal fluid in the face during a night's sleep. MATERIAL AND METHODS SUBJECTS Skin thickness was measured on 130 healthy Japanese females aged 18-83 years. All subjects had no renal or cardiac diseases. Measurements were performed on each volun- teer at three sites (the forehead, the corners of the eye [1.5 cm from the inner canthus], and the cheek [1 cm down from the top of the cheekbone]) on the face as sun-exposed areas and on the ventral side of the forearm as a poorly sun-exposed area. Before measurements, each subject washed her face with a liquid face wash (Kao Corporation, Tokyo, Japan) and stayed in a room at a constant temperature of 23øC and a relative humidity of 40% for 30 minutes for acclimation. MEASUREMENT BY B-MODE ULTRASONOGRAPHY Ultrasonography by B-mode was performed using a UX-01 ultrasonic diagnostic system (RION Co. Ltd., Tokyo, Japan) at 23øC and 40% relative humidity, as described previously (13). Briefly, the apparatus has two kinds of frequency, 15 MHz and 30 MHz, with a running width of 20 mm and 10 mm, respectively, in the direction of X, Y, and Z. The scanning mode has a B-mode with a scan time of 50 ms. The resolution ability is within 90 pm or 180 pm at 30 MHz or 15 MHz, respectively, focused at 30 mm. The velocity of ultrasound in the skin is 1530 m/sec. We measured the skin using B-mode ultrasonography under the above conditions to obtain ten ultrasonographic images per site in each subject. The B-mode ultrasonographic images obtained were printed (on thermal recording paper for Mitsubishi video copy processor SCT-K65H) using a video printer (SCT-P65 video copy processor, Mitsubishi Electric Corp, Tokyo, Japan). In preliminary experiments using A-mode ultrasonography (Dermascan A, Cortex Tech- nology, Denmark, at 1580 m/sec), we measured the skin thickness on the ventral forearms of ten females (aged 25-34 years, mean age 30 years) and compared that with the skin thickness measured using B-mode ultrasonography at the same site with dif- ferent frequencies, dynamic ranges, and gains (Figure 1). Since there was close agreement between the skin thickness measured by the above two methods (Figure 2) when B-mode

DIURNAL VARIATION AND SKIN THICKNESS 393 D B-mode image G A-mode image s Figure 1. Comparison between A- and B-modes of ultrasound image of the skin. P: membrane of the ultrasound probe. G: coupling ultrasonic gel. E: epidermal entry echo. D: dermis. S: subcutis. x: skin surface. y: interface between dermis and subcutis. • 1o0' • dynamdBdB3040rangerange 0.9' •- dynamic • •__ dynamic range 50 dB 0.8 ..... A-mode • 0.7 . ß i ß i ß ß i ' i i ß 0 4 8 12 16 20 24 28 Gain(riB) Figure 2. Comparison in the same skin site between A- and B-modes with different dynamic ranges. ultrasonography was performed in a dynamic range of 30 dB with a gain of 20 dB at 30 MHz, we used those parameters to obtain three ultrasonographic images per site in each subject in order to evaluate age-related changes in this study. MEASUREMENT OF SKIN THICKNESS On ultrasonographic images, the thickness from the skin surface to the plane showing