Portal de Periódicos da CAPES

Analysis of Circadian and Ultradian Rhythms of Skin Surface Properties of Face and Forearm of Healthy Women

2001; Elsevier BV; Volume: 117; Issue: 3 Linguagem: Inglês

10.1046/j.0022-202x.2001.01433.x

1523-1747

Isabelle Le Fur, Alain Reinberg, Sabine Lopez, Frédérique Morizot, Mohamed Mechkouri, Erwin Tschachler,

Olfactory and Sensory Function Studies

Biologic rhythms of cells and organisms are well documented and have been extensively studied at the physiologic and molecular levels. For the skin, many circadian changes have been investigated but few systematic studies comparing skin at different body sites have been reported. In this study we investigated facial and forearm skin circadian rhythms in eight healthy Caucasian women. Noninvasive methods were used to assess skin capacitance, sebum excretion, skin temperature, transepidermal water loss, and skin surface pH on fixed sites of the face and the volar forearm during a 48 h span under standardized environmental conditions. Using the cosinor or ANOVA methods, circadian rhythms could be detected for sebum excretion (face), transepidermal water loss (face and forearm), skin temperature (forearm), pH (face), and capacitance (forearm). No circadian rhythmicity was found for the other biophysical parameters. In addition to the 24 h rhythm component, rhythms with periods of 8 h were found for sebum excretion, of 8 and 12 h for transepidermal water loss (face and forearm), and of 12 h for skin temperature (forearm). Our study confirms that rhythms of skin surface parameters are readily measurable and that these rhythms differ between different sites. Furthermore, we demonstrate for the first time that, for transepidermal water loss (face and forearm), sebum excretion, and skin temperature (forearm), in addition to circadian rhythms, ultradian and/or component rhythms can be detected. Biologic rhythms of cells and organisms are well documented and have been extensively studied at the physiologic and molecular levels. For the skin, many circadian changes have been investigated but few systematic studies comparing skin at different body sites have been reported. In this study we investigated facial and forearm skin circadian rhythms in eight healthy Caucasian women. Noninvasive methods were used to assess skin capacitance, sebum excretion, skin temperature, transepidermal water loss, and skin surface pH on fixed sites of the face and the volar forearm during a 48 h span under standardized environmental conditions. Using the cosinor or ANOVA methods, circadian rhythms could be detected for sebum excretion (face), transepidermal water loss (face and forearm), skin temperature (forearm), pH (face), and capacitance (forearm). No circadian rhythmicity was found for the other biophysical parameters. In addition to the 24 h rhythm component, rhythms with periods of 8 h were found for sebum excretion, of 8 and 12 h for transepidermal water loss (face and forearm), and of 12 h for skin temperature (forearm). Our study confirms that rhythms of skin surface parameters are readily measurable and that these rhythms differ between different sites. Furthermore, we demonstrate for the first time that, for transepidermal water loss (face and forearm), sebum excretion, and skin temperature (forearm), in addition to circadian rhythms, ultradian and/or component rhythms can be detected. transepidermal water loss Biologic rhythms are defined as physiologic changes occurring over time with a reproducible waveform (Halberg and Reinberg, 1967Halberg F. Reinberg A. Rythmes circadiens et rythmes de basses frequencies en physiologie humaine.J Physiol. 1967; 59: 117-200Google Scholar). The period of rhythms encountered in biology may range from a fraction of a second, to a few hours, to about 24 h (circadian rhythms) and even longer (Scheving, 1959Scheving L.E. Mitotic activity in the human epidermis.Anat Rec. 1959; 135: 7-14Crossref PubMed Scopus (60) Google Scholar,Kahn et al., 1968Kahn G. Weinstein G.D. Frost P. Kinetics of human epidermal cell proliferation: diurnal variation.J Invest Dermatol. 1968; 50: 459-462Abstract Full Text PDF PubMed Scopus (30) Google Scholar;Gelfant et al., 1982Gelfant S. Ozawa A. Chalker D.K. Smith J.G. Circadian rhythms and differences in epidermal and in dermal cell proliferation in uninvolved and involved psoriatic skin in vivo.J Invest Dermatol. 1982; 78: 58-62Crossref PubMed Scopus (24) Google Scholar;Touitou and Haus, 1992Touitou Y. Haus E. Principles of clinical chronobiology.in: Touitou Y. Haus E. Biologic Rhythm in Clinical and Laboratory Medecine. Springer-Verlag, Berlin1992: 6-34Crossref Google Scholar). Circadian rhythms have been extensively described at different levels of physiologic organization (Touitou and Haus, 1992Touitou Y. Haus E. Principles of clinical chronobiology.in: Touitou Y. Haus E. Biologic Rhythm in Clinical and Laboratory Medecine. Springer-Verlag, Berlin1992: 6-34Crossref Google Scholar). In addition, they have attracted considerable interest in the management of cancer patients (Levi, 2000Levi F. Therapeutic implications of circadian rhythms in cancer patients.Novartis Found Symp. 2000; 227: 119-136Crossref PubMed Google Scholar) and have been associated with autoimmune and lymphoproliferative diseases in mice (Conti and Maestroni, 1998Conti A. Maestroni G.J. Melatonin rhythms in mice: role in autoimmune and lymphoproliferative diseases.Ann N Y Acad Sci. 1998; 840: 395-410Crossref PubMed Scopus (24) Google Scholar). In recent years molecular bases of circadian rhythms have been identified at the cellular level in prokaryotes and eukaryotes, and the regulation of biologic processes by circadian clocks has become a quickly expanding field of research (Dunlap, 1999Dunlap J.C. Molecular bases for circadian clocks.Cell. 1999; 10: 271-290Abstract Full Text Full Text PDF Scopus (2213) Google Scholar). Circadian rhythms of biophysical skin parameters (Burton et al., 1970Burton J.L. Cunliffe W.J. Shuster S. Circadian rhythm in sebum excretion.Br J Dermatol. 1970; 82: 497-502Crossref PubMed Scopus (34) Google Scholar;Spruit, 1971Spruit D. The diurnal variation of water vapor loss from the skin in relation to temperature.Br J Dermatol. 1971; 84: 66-70Crossref PubMed Scopus (12) Google Scholar;Timbal et al., 1972Timbal J. Colin J. Boutelier C. Guieu J.D. Evolution circadienne des températures cutanées de l'homme au repos à la neutralité thermique.Biologie, Comptes Rendus. 1972; 4-5: 512-516Google Scholar;Cotterill et al., 1973Cotterill J.A. Cunliffe W.J. Williamson B. Variation in skin surface lipid composition and sebum excretion rate with time.Acta Derm Venereol. 1973; 53: 271-274PubMed Google Scholar;Gautherie, 1973Gautherie M. Circadian rhythm in the vasomotor oscillations of skin temperature in man.Internat J Chronobiol. 1973; 1: 103-139PubMed Google Scholar;Lee et al., 1977Lee R.E. Smolensky M.H. Leach C.S. McGovern J.P. Circadian rhythms in the cutaneous reactivity to histamine and selected antigens, including phase relationship to urinary cortisol excretion.Ann Allergy. 1977; 38: 231-236PubMed Google Scholar;Marotte and Timbal, 1981Marotte H. Timbal J. Circadian rhythm of temperature in man. Comparative study with two experimental protocols.Chronobiologia. 1981; 8: 87-100PubMed Google Scholar;Verschoore et al., 1993Verschoore M. Poncet M. Krebs B. Ortonne J.P. Circadian variations in the number of actively secreting sebaceous follicles and androgen circadian rhythms.Chronobiol Internat. 1993; 10: 349-359Crossref PubMed Scopus (33) Google Scholar;Reinberg and Schmitt, 1997Reinberg A. Rythmes circadiens de la peau humaine: valeur adaptative de son organisation temporelle aux variations périodiques de l'environnement.in: Schmitt D. Biologie de la Peau Humaine. INSERM, Paris1997: 267-283Google Scholar) as well as the functional response of the skin to histamine (Reinberg et al., 1965Reinberg A. Sidi E. Ghata J. Circadian reactivity rhythms of human skin to histamine or allergen and the adrenal cycle.J Allerg. 1965; 36: 273-283Abstract Full Text PDF PubMed Scopus (53) Google Scholar,Reinberg et al., 1969Reinberg A. Zagula-Mally Z. Ghata J. Halberg F. Circadian reactivity rhythm of human skin to house dust, penicillin and histamine.J Allerg. 1969; 44: 292Abstract Full Text PDF PubMed Scopus (46) Google Scholar, Reinberg et al., 1990Reinberg A. Koulbanis C. Soudant E. Nicolaï A. Mechkouri M. Circadian changes in the size of facial skin corneocytes of healthy women.Ann Rev Chronopharmacol. 1990; 7: 331-334Google Scholar, Reinberg et al., 1996Reinberg A. Touitou Y. Soudant E. Bernard D. Bazin R. Mechkouri M. Oral contraceptives alter circadian rhythm parameters of cortisol, melatonin, blood pressure, heart rate, skin blood flow, transepidermal water loss and skin amino acids of healthy young women.Chronobiol Internat. 1996; 13: 199-211Crossref PubMed Scopus (54) Google Scholar) have been documented in the past by several investigators. Most of these studies have been performed on body skin, i.e., on the back and the volar forearm. Compared to these two regions, facial skin anatomy is more complex (Lévêque et al., 1987Lévêque J.L. Grove G. De Rigal J. Corcuff P. Kligman A.M. Saint Léger D. Biophysical characterization of dry facial skin.J Soc Cosmet Chem. 1987; 82: 171-177Google Scholar) and the face is one of the sites most exposed to environmental factors. Therefore its rhythmicity may differ from the other body areas. Circadian rhythms of facial skin have not been extensively documented as yet and the studies available are either restricted to one or two parameters (Burton et al., 1970Burton J.L. Cunliffe W.J. Shuster S. Circadian rhythm in sebum excretion.Br J Dermatol. 1970; 82: 497-502Crossref PubMed Scopus (34) Google Scholar;Timbal et al., 1972Timbal J. Colin J. Boutelier C. Guieu J.D. Evolution circadienne des températures cutanées de l'homme au repos à la neutralité thermique.Biologie, Comptes Rendus. 1972; 4-5: 512-516Google Scholar;Cotterill et al., 1973Cotterill J.A. Cunliffe W.J. Williamson B. Variation in skin surface lipid composition and sebum excretion rate with time.Acta Derm Venereol. 1973; 53: 271-274PubMed Google Scholar;Reinberg et al., 1990Reinberg A. Koulbanis C. Soudant E. Nicolaï A. Mechkouri M. Circadian changes in the size of facial skin corneocytes of healthy women.Ann Rev Chronopharmacol. 1990; 7: 331-334Google Scholar;Verschoore et al., 1993Verschoore M. Poncet M. Krebs B. Ortonne J.P. Circadian variations in the number of actively secreting sebaceous follicles and androgen circadian rhythms.Chronobiol Internat. 1993; 10: 349-359Crossref PubMed Scopus (33) Google Scholar) or model the rhythmicity after unconventional sampling schedules (Yosipovitch et al., 1998Yosipovitch G. Xiong G.L. Haus E. Sackett-Lunden L. Ashkenazi I. Maibach H.I. Time-dependent variations of the skin barrier function in humans: transepidermal water loss, stratum corneum hydration, skin surface pH and skin temperature.J Invest Dermatol. 1998; 110: 20-23https://doi.org/10.1046/j.1523-1747.1998.00069.xCrossref PubMed Scopus (210) Google Scholar). Therefore in this study we attempted to investigate simultaneously around-the-clock changes in a set of five biophysical skin parameters on the face and the forearm over a time span of 48 h. Eight healthy Caucasian women naturally menstruating, aged from 21 to 32 y (mean ±SD, 24 ± 3), were included in the study after having signed an informed consent. They had no history of ongoing or previous skin diseases. They were neither pregnant nor breast feeding, with no oral contraceptive for at least 3 mo and no medication for 15 d prior to and during the study. The influence of the menstrual cycle on skin circadian rhythms has been documented in the past (Reinberg and Smolensky, 1983Reinberg A. Smolensky M.H. Biological rhythms in Medicine. Springer Verlag, New York1983Crossref Google Scholar). Therefore the subjects were all chosen to be in the luteal phase of their menstrual cycle (28 ± 2 d cycles) during the study. All study subjects were nonsmokers. Alcohol, hot beverages, and spicy food were not permitted during the study. Subjects maintained a social and ecologic synchronization with diurnal activity with light on at 08.00 (± 1 h) and light off at midnight (± 1 h) during the 48 h study. This schedule was close to their spontaneous individual behavior. Study subjects were wearing day and night the same light cotton clothes with short sleeves leaving forearms free and were allowed to move freely in the study room. The light source consisted of fluorescent tubes Luxline-ES F 36 W/183, 120 cm (Sylvania, Erlangen, Germany), delivering only visible light (400–700 nm). Two tubes were located at the head of the bed within a distance of approximately 1.20 m from the face and 1.50 m from the volar forearm of the subject. The intensity of light reaching the skin site was about 250 lux for the face and 200 lux for the forearm. Standardized meals were served at fixed hours. Volunteers were hosted in rooms under controlled and recorded environmental conditions (temperature 20.0°C ± 0.5°C and relative humidity 53.2% ± 4.7%). Non-strenuous activities such as reading, writing, and watching TV were allowed. For standardization purposes, study subjects followed strict skin care instructions for the body and the face 1 wk prior to and during the study. In particular, they did not apply any cosmetics or make-up at least during the 12 h prior to and during the study. Moreover, they did not apply water on the investigated skin area during the study. Measurements were performed at 4 h intervals on fixed predetermined sites of the face Figure 1 and the volar forearm of subjects in recumbent position, their forearms in a horizontal position. Skin capacitance was measured using a Corneometer CM820 (Courage & Khazaka Electronic, Köln, Germany) and expressed in arbitrary units as the mean of three recordings on adjacent sites. Measurements were performed according to the EEMCO guidance for assessment of stratum corneum hydration (Berardesca, 1997Berardesca E. EEMCO guidance for the assessment of stratum corneum hydration.Skin Res Technol. 1997; 3: 126-132Crossref Scopus (418) Google Scholar). Sebum excretion was evaluated using Sebutape (Cuderm, Dallas, TX) applied on the skin surface for 1 h after cleaning the skin surface with 70° alcohol. After removing, the Sebutape was stored at -10°C until analysis with an automatic image analysis system (Quantiseb, Monaderm, Monaco). The results were expressed as percentage of Sebutape surface covered with sebum droplets, reflecting the quantity of excreted sebum in the 60 min span. Face skin temperature was recorded with a cutaneous thermometer (Differential Thermometer PT 200 from IMPO Electronics, Denmark). It was expressed in degrees Celsius as the mean of five consecutive measurements. Transepidermal water loss (TEWL), as a measurement of stratum corneum barrier function, was assessed using a Tewameter TM210C (Courage & Khazaka Electronic). At each test time a single measurement per area was performed according to the European Society of Contact Dermatitis recommendations and expressed in g per m2 h (Pinnagoda et al., 1990Pinnagoda J. Tupker R.A. Agner T. Serup J. Guidelines for transepidermal water loss (TEWL) measurement. A report from the Standardization Group of European Society of Contact Dermatitis.Contact Dermatitis. 1990; 22: 164-178Crossref PubMed Scopus (1009) Google Scholar). Skin surface pH was measured with a PHmeter PH900 (Courage & Khazaka Electronic). It was expressed as the mean of two measurements performed on two adjacent skin areas. Free salivary cortisol (µg per dl) as a marker rhythm to check subjects' synchronization (Touitou and Haus, 1992Touitou Y. Haus E. Principles of clinical chronobiology.in: Touitou Y. Haus E. Biologic Rhythm in Clinical and Laboratory Medecine. Springer-Verlag, Berlin1992: 6-34Crossref Google Scholar) was determined by enzyme-linked immunosorbent assay (BIO-Advance, Emerainville, France). To minimize interindividual differences, changes as a function of time were expressed as a percentage of the 24 h individual mean for each study subject and skin parameter. The 48 h span was selected to visualize circadian and other rhythms and to check their stability from day 1 (first 24 h span) to day 2 (second 24 h span). Day-to-day variations for each variable were tested by ANOVA. When no variation between the 2 d was found, data were pooled on a 24 h basis. Then, changes as a function of time were expressed as a percentage of the 24 h mean and displayed as a plexogram. Two complementary statistical methods were used to test 24 h (circadian), 12 h, and 8 h (ultradian) periodicities. Both cosinor and three-way ANOVA were used consecutively for the analyses. Three-way ANOVA was used to test time-dependent variations as a group phenomenon. The cosinor (Nelson et al., 1979Nelson W. Tong Y. Lee J.K. Halberg F. Methods for cosinor rhythmometry.Chronobiologia. 1979; 6: 305-323PubMed Google Scholar) was used to obtain and quantify the best-fitting cosine function approximating all data for trial periods (τ) of 24 h, 12 h, and 8 h. The least squares method was used to quantify parameters characterizing each rhythmic function. A rhythm was detected when the amplitude (half peak to trough difference) of the cosine function differed from zero with p <0.05. In these cases, the cosinor provided the acrophase (peak time location of the cosine curve) and its amplitude (half of the peak to trough difference) with their respective 95% confidence limits as well as the 24 h adjusted mean (M). When confidence limits of the acrophase were larger than ±2 h the results of the cosinor method were not considered (De Prins and Waldura, 1993De Prins J. Waldura J. Sightseeing around the single cosinor.Chronobiol Internat. 1993; 10: 395-400Crossref PubMed Scopus (43) Google Scholar) because the experimental curve is not suited to the model of a cosine function (Reinberg et al., 1998Reinberg A. Le Fur I. Tschachler E. Problems related to circadian rhythms in human skin and their validation.J Invest Dermatol. 1998; 111: 708-709https://doi.org/10.1046/j.1523-1747.1998.00336.xCrossref PubMed Scopus (12) Google Scholar). No significant day-to-day variability was found for the biophysical parameters investigated, except for skin capacitance on the face (data not shown). This allowed us to pool the data of all study subjects for the 48 h period on a 24 h basis (plexogram) except for the latter parameter. Time-dependent changes were detected by ANOVA Table I for sebum excretion (p <0.03), TEWL on the face (p <0.00005) and the forearm (p <0.00001), temperature on the forearm (p <0.00001), pH on the face (p <0.03), capacitance on the face (p <0.04) and the forearm (p <0.004), and free salivary cortisol (p <0.00001).Table IDetection of time-dependent changes of skin surface biophysical parameters in eight study subjects detected by ANOVA (with p <0.05)VariablesAnatomical sitep valueSebum excretionFace0.03TEWLFace0.00005Forearm0.00001TemperatureForearm0.00001PHFace0.03CapacitanceFace0.04Forearm0.004 Open table in a new tab To investigate the nature of the time-dependent changes we used the cosinor method. As reported previously (Haus et al., 1988Haus E. Nicolau G.Y. Lakatua D. Sackett-Lunden L. Reference values for chronopharmacology.Ann Rev Chronopharmac. 1988; 4: 333-424Google Scholar), a circadian rhythm for free salivary cortisol was detected with a peak time around 08:50 (±1 h 30 min), a nocturnal trough, and an 80% amplitude change with reference to the 24 h mean (p <0.00001) (figure not shown). This confirmed that the study subjects were synchronized. For the skin, circadian rhythms Table II were detected for sebum excretion (p <0.001), TEWL on face and forearm (p <0.0005 and p <0.03), and skin temperature on the forearm (p <0.0008). Rhythms with periods of 12 h Table III were detected for TEWL on the face (p <0.05) and the forearm (p <0.0001) and the skin temperature on the forearm (p <0.00006).Table IIDetection of circadian rhythms for skin biophysical parameters by the cosinor methodSkin variablesAnatomical siteTrial periodp valueap value of the circadian rhythm detection.AmplitudebAmplitude = half peak to trough difference. CL, confidence limits. (95% CL)AcrophasecAcrophase = peak time location expressed in hours and minutes number of study subjects = 8. (95% CL)Sebum excretionFace24 h0.00130.0 (± 24.4)13.18 (± 3.30)TEWLFace24 h0.00059.5 (± 5.9)11.38 (± 2.30)Forearm0.035.6 (± 5.3)6.00 (± 4.30)TemperatureForearm24 h0.00081.0 (± 0.8)0.48 (± 3.50)a p value of the circadian rhythm detection.b Amplitude = half peak to trough difference. CL, confidence limits.c Acrophase = peak time location expressed in hours and minutes number of study subjects = 8. Open table in a new tab Table IIIDetection of ultradian rhythms for skin biophysical parameters by the cosinor methodSkin variablesAnatomical siteTrial periodp valueap value of the circadian rhythm detection.AmplitudebAmplitude = half peak to trough difference. CL, confidence limits. (95% CL)AcrophasecAcrophase = first peak time location expressed in hours and minutes. With 12 h periods two peaks occurred in the 24 h scale at +12 h (or -12 h) from the first acrophase given in the table. With 8 h periods three peaks occurred in the 24 h scale, namely at +8 h and +16 h (or -8 h and 6 h) from the first acrophase location given in the table. Number of study subjects = 8 (95% CL) (first peak location)Sebum excretionFace8 h0.014.5 (± 1.4)5.50 (± 0.20)TemperatureForearm12 h0.000061.2 (± 0.8)5.10 (± 1.40)TEWLFace12 h0.056.0 (± 5.0)5.20 (± 1.70)8 h0.00054.5 (± 3.4)6.00 (± 0.40)Forearm12 h0.00018.9 (± 4.7)5.50 (± 1.10)8 h0.00064.1 (± 2.8)6.00 (± 0.20)CapacitanceForearm8 h0.000110.9 (± 3.8)2.10 (± 3.30)a p value of the circadian rhythm detection.b Amplitude = half peak to trough difference. CL, confidence limits.c Acrophase = first peak time location expressed in hours and minutes. With 12 h periods two peaks occurred in the 24 h scale at +12 h (or -12 h) from the first acrophase given in the table. With 8 h periods three peaks occurred in the 24 h scale, namely at +8 h and +16 h (or -8 h and 6 h) from the first acrophase location given in the table. Number of study subjects = 8 Open table in a new tab Additional rhythms with periods of 8 h were also detectable for sebum excretion on the face (p <0.01), capacitance on the forearm (p <0.0001), and TEWL on the two zones (face p <0.0005 and forearm p <0.0006) Table III. The plexograms are shown in Figure 2, Figure 3, Figure 4, Figure 5, Figure 6 and Figure 7. When a circadian rhythm was detected by cosinor analysis (Figure 3, Figure 4, Figure 5, Figure 7) the cosine functions with τ= 24 h were superimposed on the respective plexograms. The ultradian rhythms detected are not shown.Figure 3Circadian variations in sebum excretion on the forehead. Sebum excretion was determined on the forehead (see Figure 1) of the eight study subjects at 4 h intervals during 48 h. As no variations of sebum excretion measurements were found between the 2 d, data were pooled on a 24 h basis. For each subject, time point values were expressed as percentages of 24 h individual mean. Then, the mean values of these variations for the study sample (black squares, mean ± SEM) were displayed to express time-dependent changes of skin sebum excretion (plexogram). Time dependence was detected with a peak at 12:00 and a trough at 0:00. Analysis by the cosinor method detected a circadian rhythm (see also Table II) and provided the best-fitting curve that models the circadian variations for the 24 h period. This curve (dotted line) is superimposed on the corresponding plexogram. The light off period is indicated as a bold line on the time axis.View Large Image Figure ViewerDownload (PPT)Figure 4Circadian variations of TEWL on the cheeks. TEWL was measured on the left cheek (see Figure 1) of the eight study subjects at 4 h intervals during 48 h. As no variations of TEWL measurements were found between the 2 d, data were pooled on a 24 h basis. For each subject, time point values were expressed as percentages of 24 h individual mean. Then, the mean values of these variations for the study sample (black squares, mean ± SEM) were displayed to express time-dependent changes of TEWL on the cheek (plexogram). Time dependence was detected with two peaks at 8:00 and 16:00 and a trough between 20:00 and 0:00. Analysis by the cosinor method detected a circadian rhythm (see also Table II) and provided the best-fitting curve that models the circadian variations for the 24 h period. This curve (dotted line) is superimposed on the corresponding plexogram. The light off period is indicated as a bold line on the time axis.View Large Image Figure ViewerDownload (PPT)Figure 5Circadian variations of TEWL on the forearm. TEWL was measured on the volar forearm (see Figure 1) of the eight study subjects at 4 h intervals during 48 h. As no variations of TEWL measurements were found between the 2 d, data were pooled on a 24 h basis. For each subject, time point values were expressed as percentages of 24 h individual mean. Then, the mean values of these variations for the study sample (black squares, mean ± SEM) were displayed to express time-dependent changes of TEWL on the forearm (plexogram). Time dependence was detected with two peaks at 8:00 and 16:00 and two troughs at 12:00 and 0:00. Analysis by the cosinor method detected a circadian rhythm (see also Table II) and provided the best-fitting curve that models the circadian variations for the 24 h period. This curve (dotted line) is superimposed on the corresponding plexogram. The light off period is indicated as a bold line on the time axis.View Large Image Figure ViewerDownload (PPT)Figure 6Circadian variations of pH on the forehead. The pH was measured on the forehead (see Figure 1) of the eight study subjects at 4 h intervals during 48 h. As no variations of pH measurements were found between the 2 d, data were pooled on a 24 h basis. For each subject, time point values were expressed as percentages of 24 h individual mean. Then, the mean values of these variations for the study sample (black squares, mean ± SEM) were displayed to express time-dependent changes of pH (plexogram). A nocturnal trough located around 04:00 was clearly detectable. No circadian rhythm was detected by the cosinor method. The light off period is indicated as a bold line on the time axis.View Large Image Figure ViewerDownload (PPT)Figure 7Circadian variations of temperature on the volar forearm. Measurements were performed on eight subjects at 4 h intervals on the volar forearm (see Figure 1) during a sampling span of 48 h. As no variations of temperature on the volar forearm between the 2 d were found, data were pooled on a 24 h basis. For each subject, time point values were expressed as percentages of 24 h individual mean. Then, the mean values of these variations for the study sample (black squares, mean ± SEM) were displayed to express time-dependent changes of skin temperature (plexogram). Time dependence was detected with a trough at 12:00 and two peaks, a nocturnal one around 04:00 and a diurnal one around 16:00. Analysis by the cosinor method detected a circadian rhythm (see also Table II) and provided the best-fitting curve that models the circadian variations for the 24 h period. This curve (dotted line) is superimposed on the corresponding plexogram. The light off period is indicated as a bold line on the time axis.View Large Image Figure ViewerDownload (PPT) The 24 h mean values of skin capacitance for the study subjects ranged from 53 to 80 arbitrary units on the forearm (not shown). The plexogram showed three peaks Figure 2 at 12:00, 20:00, and 04:00. No circadian rhythm was detected by cosinor but an 8 h ultradian rhythmicity was detected Table III with a peak time around 02:00 (±3 h 30 min). As a measure for the sebum excretion, the 24 h mean value for the Sebutape surface covered with sebum droplets ranged from 1.0% to 18.7% for the study subjects (not shown). The plexogram (Figure 3, black squares) showed a peak at 12:00 and a trough at 0:00. By cosinor, a circadian rhythm with a peak time around 13:20 (± 3 h 30 min) was found (p <0.001) (Table II, Figure 3, dotted line). The amplitude of the cosine function was about 30% (± 24.4) with regard to the 24 h mean. The large confidence interval Table II of the acrophase indicates that the curve pattern is far from being a cosine function. On the plexogram an additional small peak occurred at 4:00 leading to a biphasic curve pattern (Figure 3, black squares). This biphasic pattern is due to the presence of a rhythm component with τ= 8 h Table III in addition to the prominent circadian rhythm. The 24 h mean TEWL of the study subjects ranged from 9.9 to 19.2 g per m2 h on the face and from 5.9 to 10.4 g per m2 h on the forearm (not shown). For the face the plexogram of the TEWL showed peaks at 8:00 and 16:00 and a trough from 20:00 to 0:00 (Figure 4, black squares). Cosinor analysis detected a peak time at about 11:40 (±2 h 30 min), which is located between the two peaks of the plexogram. The amplitude was about 9.5% (± 5.9) with regard to the 24 h mean. Again large confidence intervals for acrophase detection were found Table II. This is probably due to the fact that the 24 h curve is far from being a cosine function as it shows a two peak curve pattern. The forearm skin showed a TEWL plexogram pattern slightly different from that of the face (Figure 5, black squares) with two peaks located at 08:00 and 1