- Email: [email protected]
Facial wrinkling in postmenopausal women. Effects of smoking status and hormone replacement therapy
Maturitas 29 (1998) 75 – 86
Facial wrinkling in postmenopausal women. Effects of smoking status and hormone replacement therapy Camil Castelo-Branco a,*, Francesc Figueras a, Marı´a J Martı´nez de Osaba b, Joan A Vanrell a a
Department of Gynaecology and Obstetrics, Hospital Clı´nic i Pro6incial de Barcelona, Faculty of Medicine, Uni6ersity of Barcelona, c/ Villarroel 170, 08036 -Barcelona, Spain b Hormones Laboratory, Hospital Clı´nic i Pro6incial de Barcelona, Faculty of Medicine, Uni6ersity of Barcelona, c/ Villarroel 170, 08036 -Barcelona, Spain Received 25 September 1997; received in revised form 17 November 1997; accepted 21 November 1997
Abstract Background: There is some evidence that hormone replacement therapy may produce significant improvements in fine wrinkling, while aging skin is more frequently found in smokers. However, studies of the combined effect of a protective factor, such as HRT, and a damaging factor, such as smoking, are rare. Objecti6es: To determine in postmenopausal women the relationship between smoking status and the average number of packets of cigarettes since the subject took up smoking (packs-years) on the one hand, and facial wrinkling on the other, and to evaluate the role of hormone replacement therapy in the prevention of wrinkles in smokers and non-smokers. Methods: All subjects were recruited from our menopause clinic at Hospital Clı´nic i Provincial in Barcelona and were placed into one of three groups according to their smoking status: 215 life-long non-smokers, 306 former smokers and 209 current smokers. Smoking status, pack-years and hormone replacement were assessed by direct questioning. Facial wrinkle scores were estimated by standardized visual assessment. Results: The relative risk of moderate – severe wrinkling for current smokers compared to that for life-long non-smokers was 2.57 (confidence interval: 1.83 – 3.06; PB 0.0005). Pack-years was positively related to facial wrinkles. Life-long non-smokers receiving HRT had lower facial wrinkle scores than Life-long non-smokers who had never received HRT. HRT did not, in general, modify the facial wrinkle score in current smokers. Conclusion: Our results suggest that the risk of facial wrinkles is greater in smokers and that HRT does not diminish this risk. © 1998 Elsevier Science Ireland Ltd. Keywords: Smoking; Cigarettes; Wrinkles; Hormone replacement therapy; Menopause; Aging
* Corresponding author. 0378-5122/98/$19.00 © 1998 Elsevier Science Ireland Ltd. All rights reserved. PII S 0 3 7 8 - 5 1 2 2 ( 9 7 ) 0 0 0 8 7 - X
76
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
1. Introduction The skin, like all tissues, undergoes regressive changes with aging. The normal aging process of the skin includes gradual thinning, atrophy, dryness, skin fragility and wrinkling. Structural alterations on the molecular level lead to aged skin, while the atrophy of collagen is a major factor in the aging process. Several studies have suggested that skin collagen is affected by the loss of estrogen production by the ovaries [1 – 3] and that skin collagen content decreases in the postmenopausal years [1,4–7]. As skin collagen content declines with aging there is an increase in skin laxity and wrinkles. There is some evidence that hormone replacement therapy (HRT) may produce significant improvements in the skin’s thickness [6,8], laxity and fine wrinkling [9]. Aging skin is more frequently found in smokers. In a recent study, facial wrinkle scores were three times greater in women smokers [10]. However, studies of the combined effect of a protective factor on aging skin, such as HRT, and a damaging factor, such as smoking, are scarce. We evaluated cigarette smoking and facial wrinkling in 730 postmenopausal women while taking into consideration other potential variables including age, exposure to the sun, alcohol consumption, body mass index and hormone replacement therapy use.
2. Subjects and methods
2.1. Subjects Before being included in the study, all patients underwent a medical history review and physical examination, including wrinkles’ evaluation and gynecologic examination, pelvic ultrasound, and bone absorptiometry. All subjects were recruited from our menopause clinic at Hospital Clı´nic i Provincial in Barcelona between 1994 and 1996. Based on the data collected at the first visit, all subjects were arranged by age in four 5-year age groups between 40 and 60, by smoking status into three groups: life-long non-smokers, former (] 24 months) and current smokers and by hormone replacement into two groups: those who had never
received HRT and current recipients (\ 12 months use). Based on differences in wrinkling between smokers and life-long non-smokers [11], between lifelong non-smokers, former and current smokers [10] and between local HRT and placebo [9] previously reported and our hypothesis that HRT would have a protective effect on wrinkling, we estimated that we would need 14 life-long non-smokers, 14 current smokers and 22 former smokers in each of the eight age-HRT strata [12]. As there is little information describing the effect of previous HRT on present skin conditions and, moreover, since the inclusion of this group would have greatly increased the number of subjects required, former HRT-users were not considered in this study.
2.2. Methods Smoking status, pack-years of smoking and hormone replacement were obtained from medical records or assessed by direct questioning. Facial wrinkle scores (WS) were estimated by standardized visual assessment. During their first visit, and after washing and removing any make-up from subjects, we evaluated three facial areas and scored each for facial wrinkling. The areas were: first, the crow’s feet around the right eye, i.e. the area compressed radially from 0.5 cm medial to lateral canthus of the right eye to the hairline; second, the forehead area, i.e. the area limited laterally by the hairlines and contained between brows and the hairline; and third, the upper perioral area, i.e. the skin bounded laterally by the lines from the corner of the mouth to the lateral nasolabial fold and contained between the line below the nose and the upper lip. The number, depth (superficial, medium, deep) and length (cm) of all the wrinkles in each area were recorded. The wrinkle score (WS) was estimated by multiplying the depth of each wrinkle by its length and adding the product up to a maximum of five wrinkles in each area. The wrinkle scoring system is given in Table 1. The wrinkle patterns were defined as none (basically no wrinkles) when the score was 5 5, mild if the WS was \ 5 but B 30, moderate when the WS was \30 but B90 and severe when WS was ] 90 (maximum score: 150). All subjects were
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
evaluated by a single observer (CCB) who was blinded with regard to women’s smoking and Table 1 Wrinkles’ score Item
Category
Value
Number
1 2 3 4 (])5 Shallow Medium Deep (5)0.5 0.5–1 1.0–1.5 1.5–2 (])2
1 2 3 4 5 1 3 6 1 2 3 4 5
Depth
Length (cm)
HRT status. To determine the level of exposure to tobacco smoke, each subject was asked how many packs of cigarettes per day she had smoked over each 5-year period of her life since taking up smoking. These 5-year periods were then summed to give a ‘pack-year’ smoking history (i.e. 1 pack/day for 20 years would be a 20 pack-year history). In addition, all subjects were questioned about the nicotine and tar content of their cigarettes. Former smokers were considered as those who had not smoked during the 24 months prior to inclusion in the study. In addition, body mass index (BMI), sun exposure (mean hours/day since the age of 18) and alcohol consumption (g/day) were recorded.
2.3. Statistical analysis
Wrinkles’ score (WS) was estimated according to the formula: WS =W a+b+c/3 where
W a+b+c =%a(1−5)(d.l)+%b(1−5)(d.l)+%c(1−5)(d.l) by multiplying the value of depth (d) by the value of length (l) of each wrinkle and adding this product for all wrinkles with a maximum of five ((1-5)) in the three facial areas: right crow’s foot (a), forehead (b) and upper perioral area (c). Table 2 Distribution of the study subjects by age, HRT use and smoking status Smoking status
77
Total (n)
Never (n)
Former (n)
Current (n)
HRT 40–44 45–49 50–54 55–60
107 24 23 30 30
150 36 44 34 36
103 33 23 23 24
360 93 90 87 90
No HRT 40–44 45–49 50–54 55–60
108 24 21 30 33
156 26 34 43 53
106 35 29 16 26
370 85 84 89 112
Total
215
306
209
730
Statistical analysis were performed using the Statistical Package for the Social Sciences (SPSS for Windows). Kolmogorov’s test was used to confirm the normal distribution of the sample (Skewness: 1.76; Kurtosis: 2.95). Multiple linear regression and multiple analysis of variance for repeated measures (MANOVA) were used to model bivariate relationships, to test for additive effects, and to control for possible confounding variables such as age, BMI and sun exposure. Correlations between two continuous variables were assessed with Spearman’s rank test, and for comparison of the means MANOVA was used. Confidence intervals for the correlations and prediction intervals were calculated to evaluate the accuracy of the regressions, and finally, clinical assessment of therapeutic efficacy was analyzed with the x 2-squared test using Pearson, Likelihood Ratio or Mantel–Haenszel test for linear association where appropriate. A P-value below 0.05 was considered to indicate statistical significance.
3. Results The distribution of the patients by age, HRT use and smoking status is shown in Table 2 and plotted as linear regression in Figs. 1 and 2. The
78
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
Fig. 1. Relationship between Wrinkle Score (WS) and age according to smoking status (SM) for current smokers (closed circles; solid line), former smokers (asterisk; dotted line) and never smokers (open circles; dotted line). Data from 730 women are shown. Note the significant shift of the WS for current smokers vs former and never smokers.
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
79
Fig. 2. Relationship between Wrinkle Score (WS) and age according to hormone replacement therapy (HRT) for users (closed circles; dotted line), never users (triangle; dotted line) and total population (solid line). Data from 730 women are shown. Although not significant, mean WS was lower for HRT users across all ages.
80
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
Fig. 3. Percentage of women with moderate–severe wrinkle patterns (WS \ 30) by HRT use according to age groups and smoking status (never smokers in black columns, former and current in patterned columns).
subjects were arranged into three groups according to their smoking habits: 215 life-long nonsmokers, 306 former smokers and 209 current smokers and into two HRT-related groups: 360 users and 370 non-users. The percentage of use of low-level nicotine cigarettes was similar among HRT users (3.3%) and non-users (3.5%).
The relative risk (RR) of moderate–severe wrinkling for current smokers versus life-long non-smokers was 2.57 (CI: 1.83–3.6, P B 0.0005) and 1.42 (CI: 1.0–2.04, P B 0.05) versus former smokers. The percentages of subjects according to smoking status and HRT use who presented moderate
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
81
Table 3 Effect of HRT and smoking status on wrinkles Smoking status Never
Former
Current
n
WS
n
WS
n
WS
HRT users No HRT
107 (15) 108 (22)
23.79 12.4 30.69 17.2
150 (30) 156 (43)
27.9 9 14.2 30.9 914.6
103 (41) 106 (49)
38.9 9 23.0* 43.5 924.6**
Total
215 (37)
27.29 15.3
306 (73)
29.5 914.5
209 (90)
41.2 923.9*
n, number; WS, Wrinkles’ score. In parentheses number of patients with WS\30. WS are expressed as mean 9S.D. Although differences were significant between current and never and former smokers in both HRT-related groups — users and never users— when comparing HRT users with HRT never users, no differences were observed. * PB0.0002. ** PB0.005 current vs never and former smokers. Table 4 Relative risk for moderate or severe facial wrinkling (WS\30) adjusted for age and HRT for two models: (a) smokers — current or former — compared with never smokers and (b) active smokers compared with no present smokers — never or former Age
HRT status HRT users
HRT never users
n
RR
CI
n
RR
CI
(a) 40–44 45–49 50–54 55–60
253 69 67 57 60
2.4 1.8 1.6 2.7 4.1
1.3–4.4*** 0.5–5.9 0.4–6.2 0.7–10.2 1.4–12.1
262 61 63 59 79
2.1 1.5 8.1 1.2 2.0
1.2 – 3.6*** 0.6 – 3.9 1.7 – 37.8 0.4 – 3.4 0.7 – 5.9
(b) 40–44 45–49 50–54 55–60
107 24 25 29 29
3.2 4.1 3.5 1.9 1.9
1.2–4.8*** 1.0–15.7** 1.2–10.5*** 1.0–3.7* 1.2–3.0*
108 25 26 25 32
2.8 3.0 4.7 1.5 1.6
1.1 – 3.9*** 1.0 – 8.8** 1.7 – 12.7*** 0.9 – 2.5 1.1 – 2.2*
RR, relative risk; CI: 95% confidence interval. * PB0.05. ** PB0.01. *** PB0.005.
or severe WS are shown in Fig. 3. As expected the percentage of subjects who presented moderate – severe WS increased with age in all groups (Figs. 1 – 3). However, a higher proportion of current smokers had moderate to severe WS than life-long non-smokers. Across all age groups, the proportion of former smokers who had moderate – severe WS was higher than life-long non-smokers but lower than for current smokers. HRT did not significantly modify the facial wrinkle score in
current smokers and when comparing HRT users with non-users the only difference observed was among the life-long non-smoker subjects (Table 3). The RR for moderate–severe wrinkling for current smokers adjusted for age was 1.48 (CI: 1.21– 1.80; PB 0.0001) in the HRT non-user group and 1.44 (CI: 1.21–1.72; P B 0.00001) in the HRT user group. The RR of moderate–severe wrinkling for former smokers was 1.09 (CI: 0.95–1.25) in the
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
82
Table 5 Multiple linear regression for wrinkle score (dependent variable) adjusted for packs/year. The effect of smoking on wrinkles’ score was independent of age and HRT Variable
B
SE for B
Beta
T
P-value
Age HRT SM PY (Constant)
1.150746 −4.373457 −2.333396 0.994492 −31.746534
0.093797 1.099150 1.066535 0.083972 4.789901
0.361805 −0.116400 −0.094663 0.514296
12.268 −3.979 −2.188 11.843 −6.628
0.0000 0.0001 0.0290 0.0000 0.0000
Multiple R: 0.61614; R squared: 0.37963; F: 110.91; PB0.0000. SE, standard error; HRT, Hormone Replacement Therapy; SM, Smoking Status; PY: packs/year.
HRT non-user group and 1.08 (CI: 0.97 – 1.21) in the HRT user group. The RR of moderate – severe wrinkling for smokers including both former and current smokers was 1.72 (CI: 1.14 – 2.59; P B0.005) in the HRT non-user group and 2.14 (CI: 1.26 – 3.63; PB 0.002) in the HRT user group (Table 4). Finally, the effect of smoking on wrinkle score was found to be independent of age and HRT (Table 5). Pack-years was positively correlated to facial wrinkles (Fig. 4). Postmenopausal women with pack-year scores over 10 were more prone to have a WS\ 30 (moderate and severe). In analyses adjusted for age, the RR associated with more than 20 pack-years of exposure compared with life-long non-smokers was 1.48 (CI: 1.13 – 1.93; P B0.005) among the HRT users and 1.71 (CI: 1.25 –2.34; PB0.00002) among the HRT nonusers. The proportion of life-long non-smokers on HRT who had a moderate to severe facial WS compared with life-long non-smokers who were HRT non-users was clearly lower (Fig. 3). This difference was more apparent in women over 50 years of age. HRT did not, in general, modify the facial wrinkle score in current smokers, only the group of subjects who smoke or smoked less than 10 pack-years illustrated a protective effect of HRT on WS (Fig. 4). Other variables that may affect facial WS were also studied (Table 6). Sun exposure of 4 h/day or more was associated with higher values of WS (r: 0.579, P B0.01) and with an increased risk of wrinkling. Alcohol consumption ]40 g/day was
also associated with higher values of WS (r: 0.371, PB 0.05). Finally, BMI was inversely related to facial wrinkling (r: − 0.432, PB 0.05).
4. Discussion Great interest has been aroused in therapies that might delay the process of aging. Human aging is a dynamic, unchangeable phenomenon which affects all systems in the body and is commonly perceived as an intrinsic process within a community affecting all of its members. Among the most characteristic aging changes are those undergone by the skin. Skin aging results from an association of chronologic, hormonal (i.e. menopause) and environmental (i.e. tobacco, sun exposure, alcohol, etc.) factors. Classically, chronologic aging includes those cutaneous changes that occur in areas not exposed to the sun, such as the buttocks, and are observed in both men and women. The regressive changes which the skin undergoes with age, such as loss of elasticity, epidermal thickness, and elastic degeneration, may account for the appearance of clinical examples, which include soft tissue sagging and wrinkles. It has been demonstrated that skin collagen is affected by hypoestrogenism [1–3] and that the amount of skin collagen declines in the years following the menopause [1,4–7]. Skin thickness, which was also considered in some of these studies, decreased at a similar rate after the menopause [4–6]. Moreover, the decrease in the amount of skin collagen correlates well with the
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
83
Fig. 4. Relationship between Wrinkle Score (WS) and packs of cigarettes/year according to hormone replacement therapy (HRT) for users (closed circles; dotted line), never users (open circles; dotted line) and total population (solid line). Data from 730 women are shown. Note the significant increase of the WS after 10 packs/year consumption.
reduction in skin thickness observed in the years following castration [8]. Additionally, one of the most significant characteristics of skin biomechanics related to collagen is tensile strength, and several studies have detected a decrease in this parameter with aging [13,14]. Based on these findings, efforts have been made to slow down the degenerative changes in aging skin with oestrogen therapy. As far back as 1946, Goldzieher [15] observed that oestrogens had a regenerative effect on the senile epidermis. This early finding is supported by recent studies which have demonstrated
that oestrogen replacement therapy may increase skin thickness [6,8,16] and skin collagen content [1,4,5]. Even local oestriol may improve the structure of elastic fibres [17]. It is well known that aging skin is more frequently apparent in smokers. Here, we found that current smokers have a RR for facial wrinkling three times that of life-long non-smokers and almost twice that of former smokers. Our results are similar to those reported elsewhere [10,18,19]. This study sought to determine whether HRT neutralized the effects of smoking on the skin.
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
84
Table 6 Relative risk (RR) for moderate or severe facial wrinkling (WS\30) adjusted for age by chosen risk factors HRT status HRT users Variable Agea 40–44 45–49 50–54 55–60 Smoking status Never smoker Former smoker Current smoker Pack/years Never smoker B5 5–10 10–20 \20 Sun exposure B2 h/day 2–4 h/day \4 h/day BMI Lower third Middle third Upper third Alcohol consumption Non-drinker B20 g/day 20–40 g/day \40 g/day
HRT never users RR
CI
n
RR
CI
86 87 89 98
1.0 1.03 1.28 1.53
— 0.93–1.15 1.10–1.49* 1.28–1.83***
85 84 89 112
1.0 2.2 3.1 4.2
— 1.6 – 4.5 2.5 – 6.0** 2.2 – 7.1***
107 150 103
1.0 1.08 1.44
— 0.97–1.21 1.21–1.72**
108 156 106
1.0 1.09 1.48
— 0.95 – 1.25 1.21 – 1.80**
107 25 64 88 76
1.0 0.94 1.03 1.23 1.53
— 0.82–1.08 0.91–1.17 1.05–1.44* 1.24–1.89**
108 36 46 91 89
1.0 0.89 1.14 1.27 1.44
— 0.77 – 1.04 0.92 – 1.42 1.06 – 1.53* 1.17 – 1.79**
72 125 166
1.0 1.36 2.09
— 0.95–2.94* 1.41–3.28**
91 170 106
1.0 1.72 2.26
— 0.87 – 3.19* 1.70 – 4.24**
129 131 133
1.0 0.78 0.64
— 0.45–1.30++ 0.47–1.26+
102 135 130
1.0 0.81 0.69
— 0.50 – 1.59++ 0.47 – 1.18+
35 189 111 28
1.0 0.90 1.15 1.27
— 0.69–1.87 0.85–2.36 1.10–2.90++
45 206 94 22
1.0 0.87 1.20 1.35
— 0.67 – 1.75 0.77 – 1.99 1.12 – 2.97++
n
CI: 95% confidence interval. Not adjusted. ++ PB0.05. + PB0.01. * PB0.001. ** PB0.0001. *** PB0.00000. a
Although this study included a large number of subjects, each of which were evaluated by a single observer, we cannot rule out altogether the possibility of a bias in the scoring for smokers. However, the number of subjects was sufficient to evaluate the effect of smoking and HRT on the one hand with facial wrinkling by age groups on the other, to consider current and former smokers separately and to control for several variables
such as age, pack-years, sun exposure, alcohol consumption and body mass. Our study like others conducted in the same field [10,18,19], still presents certain weaknesses, including the lack of biochemical verification of smoking status, alcohol consumption or HRT use; the fact that several potential confounding factors e.g. the use of sunblocks, sunglasses or other protective products, were not recorded; the possibility of bias in sun
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
exposure across groups; the fact that the wrinkle score was an estimate based only on visual assessment; and finally, the possibility that smoking, use of HRT, and severity of facial wrinkling may be inter-related through other unstudied variables. However, since data were recorded individually in a confidential interview we do not expect the reports of cigarette or alcohol use by subjects to be incorrect. The effect of potential confounding factors, such as skin protective products, might have affected our results, but only if they were associated with smoking status, HRT and wrinkling. Clearly, ultrasonography for measuring skin thickness or profilometry for cutaneous microrelief, as described by Creidi et al. [9], would have been more objective than visual assessment; however, the size of our population meant that the implementation of the two procedures would have been highly complicated. Among smokers (current and former) the RR for wrinkling appeared to increase after 10 packyears. This results, shown in Fig. 4, is in agreement with data from Ernster et al. who found an increased risk for wrinkles in women after the same number of pack-years [10]. We can only hypothesize as to the mechanisms that might be involved in skin aging and facial wrinkling and to their relationship with the use of HRT and the habit of smoking. It has been reported that the ground substance composed mainly of glycosaminoglycans and glycoproteins diminishes in aged skin [20]. Hyaluronic acid is an important constituent and is responsible for the normal turgor of the dermis because of its extraordinary water-holding capacity. The ground substance also contributes to the rheological properties of the skin and may be affected by the components of tobacco. Many of the acute vascular effects of tobacco have been attributed to nicotine [21]; however, tobacco smoke has a great variety of components that may influence directly and indirectly the ground substance. This include tarry and phenolic products, volatile acids and gaseous particles that may account for a chronic hypoxemia of the skin. Moreover, cigarette smoke causes a decrease in the capillary blood flow in the skin [21,22] which may be a causative factor in wrinkling as it affects hyaluronic acid and gly-
85
cosaminoglycan synthesis. To corroborate this hypothesis, further studies are needed. Efforts to reverse the effects of aging skin have been attempted since ancient times, and because of its huge psychological impact, there is much demand for effective treatment. Many well-constructed trials confirm the clinical efficacy of modern therapies with oestrogens for improving the features of aged skin [1,4–6,8,9,15–17]. The effect of estrogens can be seen by increasing glycosaminoglycans and hyaluronic acid concentration, and thereby increasing the hydration of collagenous tissue of the dermis. Despite many claims, there have been no adequate trials documenting the efficacy of fish cartilage polysaccharide extracts or alpha-hydroxy acids, only topical tretionin have been reported to reverse fine wrinkling and mild to moderate hyperpigmentation [23,24]. Although tretionin compounds may be helpful, they also present inconveniences such as skin intolerance, discomfort and a considerable expense [23]. Careful patient selection with consideration of alternative treatments such as chemical peeling [25], dermabrasion, implants [26], and surgery is important in the successful management of skin aging. Finally, our results indicate an accelerated skin aging in current smokers and an improvement in WS only in the life-long non-smoker group, as a result of HRT. Thus, rather than promoting HRT as the solution, smokers need to be made aware of the increased risks of skin aging as a result of their habit. References [1] Castelo-Branco C, Dura´n M, Gonza´lez-Merlo J. Skin collagen changes related to age and hormone replacement therapy. Maturitas 1992;15:113 – 9. [2] Brincat M, Kabalan S, Studd JWW, Moniz CF, de Trafford J, Montgomery J. A study of the decrease of skin collagen content, skin thickness and bone mass in the postmenopausal woman. Obstet Gynecol 1987;70:840 – 5. [3] Brincat M, Moniz CF, Studd JWW, Darby AJ, Magos A, Cooper D. Sex hormones and skin collagen content in postmenopausal women. Br Med J 1983;287:1337 – 8. [4] Brincat M, Versi E, Moniz CJ, Magos A, de Trafford J, Studd JWW. Skin collagen changes in postmenopausal women receiving different regimens of estrogen therapy. Obstet Gynecol 1987;70:123 – 7.
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
86
[5] Brincat M, Moniz CF, Kabalan S, Versi E, O’Dowd T, Magos AL, Montgomery J, Studd JWW. Decline in skin collagen content and metacarpial index after the menopause and its prevention with sex hormone replacement. Br J Obstet Gynecol 1987;94(2):126–9. [6] Brincat M, Moniz CF, Studd JWW, Darby AJ, Magos A, Emburey G, Versi E. Long-term effects of the menopause and sex hormones on skin thickness. Br J Obstet Gynecol 1985;92:256 – 9. [7] Castelo-Branco C, Pons F, Gratacos E, Fortuny A, Vanrell JA, Gonza´lez-Merlo J. Relationship between skin collagen and bone changes during aging. Maturitas 1994;18:199 – 206. [8] Punnonen R. Effect of castration and peroral oestrogen therapy on the skin. Acta Obstet Gynecol Scand 1977;21(Suppl):1– 44. [9] Creidi P, Faivre B, Agache P, Richard E, Haudiquet V, Sauvanet JP. Effect of a conjugated oestrogen cream on ageing facial skin. A comparative study with a placebo cream. Maturitas 1994;19:211–23. [10] Ernster VL, Grady D, Miike R, Black D, Selby J, Kerlikowske K. Facial wrinkling in men and women by smoking status. Am J Public Health 1995;85:78–82. [11] Daniell HW. Smoker’s wrinkles. Ann Intern Med 1971;75:873 – 80. [12] Pardell H, Cobo E, Canela J. Manual de Bioestad’stica. Barcelona: Masson Editorial, 1986:183–185. [13] Vogel HG. Influence of maturation and age on mechanical and biomechanical parameters of connective tissue of various organs in the rat. Connect Tissue Res 1978;6:161 – 6. [14] Vogel HG. Attempts to compare ‘in vivo’ and ‘in vitro’ measurements of mechanical properties in rat skin. Bioeng Skin 1981;3:39 –46. [15] Goldzieher MA. The effects of oestrogens on the senile skin. J Gerontol 1946;1:196–201.
.
[16] Punnonnen R, Vilska S, Rauramo L. Skinfold thickness and long-term postmenopausal hormone therapy. Maturitas 1984;5:259 – 62. [17] Punnonnen R, Vaajalahti P, Teisala K. Local oestriol treatment improves the structure of elastic fibers in the skin of postmenopausal women. Ann Chirurg Gynaecol 1987;76(Suppl 202):39 – 41. [18] Kadunce DP, Burr R, Gress R, et al. Cigarette smoking: risk factor for premature facial wrinkling. Ann Intern Med 1991;114:840 – 4. [19] Model D. Smokers’ faces; who are the smokers?. Br Med J 1985;291:1760– 2. [20] Fleishmajer R, Perlish JJ, Bashey RI. Aging of human dermis. In: Pobert CL, editor. Frontiers of Matrix Biology. Basel: Karger, 1973:90 – 106. [21] Reus WF, Robson MC, Zachary L, Heggers JP. Acute effects of tobacco smoking on blood flow in the cutaneous micro-circulation. Br J Plast Surg 1984;37:213 – 5. [22] Richardson D. Effects of tobacco smoke inhalation on capillary blood flow in human skin. Arch Environ Health 1987;42:19 – 25. [23] Noble S, Wagstaff AJ. Tretinoin. A review of its pharmacological properties and clinical efficacy in the topical treatment of photodamaged skin. Drugs Aging 1995;6(6):479 – 96. [24] Helander SD. Treatment of photoaged skin. Efficacy, tolerability and costs of available agents. Drugs Aging 1996;8(1):12 – 6. [25] Humphreys TR, Werth V, Dzubow L, Kligman A. Treatment of photodamaged skin with trichloroacetic acid and topical tretinoin. J Am Acad Dermatol 1996;34:638 – 44. [26] Lemperle G, Hazan-Gauthier N, Lemperle M. PMMA microspheres (Artecoll) for skin and soft-tissue augmentation. Part II: Clinical investigations. Plast Reconstr Surg 1995;96:627 – 34.
.
Facial wrinkling in postmenopausal women. Effects of smoking status and hormone replacement therapy Camil Castelo-Branco a,*, Francesc Figueras a, Marı´a J Martı´nez de Osaba b, Joan A Vanrell a a
Department of Gynaecology and Obstetrics, Hospital Clı´nic i Pro6incial de Barcelona, Faculty of Medicine, Uni6ersity of Barcelona, c/ Villarroel 170, 08036 -Barcelona, Spain b Hormones Laboratory, Hospital Clı´nic i Pro6incial de Barcelona, Faculty of Medicine, Uni6ersity of Barcelona, c/ Villarroel 170, 08036 -Barcelona, Spain Received 25 September 1997; received in revised form 17 November 1997; accepted 21 November 1997
Abstract Background: There is some evidence that hormone replacement therapy may produce significant improvements in fine wrinkling, while aging skin is more frequently found in smokers. However, studies of the combined effect of a protective factor, such as HRT, and a damaging factor, such as smoking, are rare. Objecti6es: To determine in postmenopausal women the relationship between smoking status and the average number of packets of cigarettes since the subject took up smoking (packs-years) on the one hand, and facial wrinkling on the other, and to evaluate the role of hormone replacement therapy in the prevention of wrinkles in smokers and non-smokers. Methods: All subjects were recruited from our menopause clinic at Hospital Clı´nic i Provincial in Barcelona and were placed into one of three groups according to their smoking status: 215 life-long non-smokers, 306 former smokers and 209 current smokers. Smoking status, pack-years and hormone replacement were assessed by direct questioning. Facial wrinkle scores were estimated by standardized visual assessment. Results: The relative risk of moderate – severe wrinkling for current smokers compared to that for life-long non-smokers was 2.57 (confidence interval: 1.83 – 3.06; PB 0.0005). Pack-years was positively related to facial wrinkles. Life-long non-smokers receiving HRT had lower facial wrinkle scores than Life-long non-smokers who had never received HRT. HRT did not, in general, modify the facial wrinkle score in current smokers. Conclusion: Our results suggest that the risk of facial wrinkles is greater in smokers and that HRT does not diminish this risk. © 1998 Elsevier Science Ireland Ltd. Keywords: Smoking; Cigarettes; Wrinkles; Hormone replacement therapy; Menopause; Aging
* Corresponding author. 0378-5122/98/$19.00 © 1998 Elsevier Science Ireland Ltd. All rights reserved. PII S 0 3 7 8 - 5 1 2 2 ( 9 7 ) 0 0 0 8 7 - X
76
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
1. Introduction The skin, like all tissues, undergoes regressive changes with aging. The normal aging process of the skin includes gradual thinning, atrophy, dryness, skin fragility and wrinkling. Structural alterations on the molecular level lead to aged skin, while the atrophy of collagen is a major factor in the aging process. Several studies have suggested that skin collagen is affected by the loss of estrogen production by the ovaries [1 – 3] and that skin collagen content decreases in the postmenopausal years [1,4–7]. As skin collagen content declines with aging there is an increase in skin laxity and wrinkles. There is some evidence that hormone replacement therapy (HRT) may produce significant improvements in the skin’s thickness [6,8], laxity and fine wrinkling [9]. Aging skin is more frequently found in smokers. In a recent study, facial wrinkle scores were three times greater in women smokers [10]. However, studies of the combined effect of a protective factor on aging skin, such as HRT, and a damaging factor, such as smoking, are scarce. We evaluated cigarette smoking and facial wrinkling in 730 postmenopausal women while taking into consideration other potential variables including age, exposure to the sun, alcohol consumption, body mass index and hormone replacement therapy use.
2. Subjects and methods
2.1. Subjects Before being included in the study, all patients underwent a medical history review and physical examination, including wrinkles’ evaluation and gynecologic examination, pelvic ultrasound, and bone absorptiometry. All subjects were recruited from our menopause clinic at Hospital Clı´nic i Provincial in Barcelona between 1994 and 1996. Based on the data collected at the first visit, all subjects were arranged by age in four 5-year age groups between 40 and 60, by smoking status into three groups: life-long non-smokers, former (] 24 months) and current smokers and by hormone replacement into two groups: those who had never
received HRT and current recipients (\ 12 months use). Based on differences in wrinkling between smokers and life-long non-smokers [11], between lifelong non-smokers, former and current smokers [10] and between local HRT and placebo [9] previously reported and our hypothesis that HRT would have a protective effect on wrinkling, we estimated that we would need 14 life-long non-smokers, 14 current smokers and 22 former smokers in each of the eight age-HRT strata [12]. As there is little information describing the effect of previous HRT on present skin conditions and, moreover, since the inclusion of this group would have greatly increased the number of subjects required, former HRT-users were not considered in this study.
2.2. Methods Smoking status, pack-years of smoking and hormone replacement were obtained from medical records or assessed by direct questioning. Facial wrinkle scores (WS) were estimated by standardized visual assessment. During their first visit, and after washing and removing any make-up from subjects, we evaluated three facial areas and scored each for facial wrinkling. The areas were: first, the crow’s feet around the right eye, i.e. the area compressed radially from 0.5 cm medial to lateral canthus of the right eye to the hairline; second, the forehead area, i.e. the area limited laterally by the hairlines and contained between brows and the hairline; and third, the upper perioral area, i.e. the skin bounded laterally by the lines from the corner of the mouth to the lateral nasolabial fold and contained between the line below the nose and the upper lip. The number, depth (superficial, medium, deep) and length (cm) of all the wrinkles in each area were recorded. The wrinkle score (WS) was estimated by multiplying the depth of each wrinkle by its length and adding the product up to a maximum of five wrinkles in each area. The wrinkle scoring system is given in Table 1. The wrinkle patterns were defined as none (basically no wrinkles) when the score was 5 5, mild if the WS was \ 5 but B 30, moderate when the WS was \30 but B90 and severe when WS was ] 90 (maximum score: 150). All subjects were
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
evaluated by a single observer (CCB) who was blinded with regard to women’s smoking and Table 1 Wrinkles’ score Item
Category
Value
Number
1 2 3 4 (])5 Shallow Medium Deep (5)0.5 0.5–1 1.0–1.5 1.5–2 (])2
1 2 3 4 5 1 3 6 1 2 3 4 5
Depth
Length (cm)
HRT status. To determine the level of exposure to tobacco smoke, each subject was asked how many packs of cigarettes per day she had smoked over each 5-year period of her life since taking up smoking. These 5-year periods were then summed to give a ‘pack-year’ smoking history (i.e. 1 pack/day for 20 years would be a 20 pack-year history). In addition, all subjects were questioned about the nicotine and tar content of their cigarettes. Former smokers were considered as those who had not smoked during the 24 months prior to inclusion in the study. In addition, body mass index (BMI), sun exposure (mean hours/day since the age of 18) and alcohol consumption (g/day) were recorded.
2.3. Statistical analysis
Wrinkles’ score (WS) was estimated according to the formula: WS =W a+b+c/3 where
W a+b+c =%a(1−5)(d.l)+%b(1−5)(d.l)+%c(1−5)(d.l) by multiplying the value of depth (d) by the value of length (l) of each wrinkle and adding this product for all wrinkles with a maximum of five ((1-5)) in the three facial areas: right crow’s foot (a), forehead (b) and upper perioral area (c). Table 2 Distribution of the study subjects by age, HRT use and smoking status Smoking status
77
Total (n)
Never (n)
Former (n)
Current (n)
HRT 40–44 45–49 50–54 55–60
107 24 23 30 30
150 36 44 34 36
103 33 23 23 24
360 93 90 87 90
No HRT 40–44 45–49 50–54 55–60
108 24 21 30 33
156 26 34 43 53
106 35 29 16 26
370 85 84 89 112
Total
215
306
209
730
Statistical analysis were performed using the Statistical Package for the Social Sciences (SPSS for Windows). Kolmogorov’s test was used to confirm the normal distribution of the sample (Skewness: 1.76; Kurtosis: 2.95). Multiple linear regression and multiple analysis of variance for repeated measures (MANOVA) were used to model bivariate relationships, to test for additive effects, and to control for possible confounding variables such as age, BMI and sun exposure. Correlations between two continuous variables were assessed with Spearman’s rank test, and for comparison of the means MANOVA was used. Confidence intervals for the correlations and prediction intervals were calculated to evaluate the accuracy of the regressions, and finally, clinical assessment of therapeutic efficacy was analyzed with the x 2-squared test using Pearson, Likelihood Ratio or Mantel–Haenszel test for linear association where appropriate. A P-value below 0.05 was considered to indicate statistical significance.
3. Results The distribution of the patients by age, HRT use and smoking status is shown in Table 2 and plotted as linear regression in Figs. 1 and 2. The
78
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
Fig. 1. Relationship between Wrinkle Score (WS) and age according to smoking status (SM) for current smokers (closed circles; solid line), former smokers (asterisk; dotted line) and never smokers (open circles; dotted line). Data from 730 women are shown. Note the significant shift of the WS for current smokers vs former and never smokers.
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
79
Fig. 2. Relationship between Wrinkle Score (WS) and age according to hormone replacement therapy (HRT) for users (closed circles; dotted line), never users (triangle; dotted line) and total population (solid line). Data from 730 women are shown. Although not significant, mean WS was lower for HRT users across all ages.
80
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
Fig. 3. Percentage of women with moderate–severe wrinkle patterns (WS \ 30) by HRT use according to age groups and smoking status (never smokers in black columns, former and current in patterned columns).
subjects were arranged into three groups according to their smoking habits: 215 life-long nonsmokers, 306 former smokers and 209 current smokers and into two HRT-related groups: 360 users and 370 non-users. The percentage of use of low-level nicotine cigarettes was similar among HRT users (3.3%) and non-users (3.5%).
The relative risk (RR) of moderate–severe wrinkling for current smokers versus life-long non-smokers was 2.57 (CI: 1.83–3.6, P B 0.0005) and 1.42 (CI: 1.0–2.04, P B 0.05) versus former smokers. The percentages of subjects according to smoking status and HRT use who presented moderate
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
81
Table 3 Effect of HRT and smoking status on wrinkles Smoking status Never
Former
Current
n
WS
n
WS
n
WS
HRT users No HRT
107 (15) 108 (22)
23.79 12.4 30.69 17.2
150 (30) 156 (43)
27.9 9 14.2 30.9 914.6
103 (41) 106 (49)
38.9 9 23.0* 43.5 924.6**
Total
215 (37)
27.29 15.3
306 (73)
29.5 914.5
209 (90)
41.2 923.9*
n, number; WS, Wrinkles’ score. In parentheses number of patients with WS\30. WS are expressed as mean 9S.D. Although differences were significant between current and never and former smokers in both HRT-related groups — users and never users— when comparing HRT users with HRT never users, no differences were observed. * PB0.0002. ** PB0.005 current vs never and former smokers. Table 4 Relative risk for moderate or severe facial wrinkling (WS\30) adjusted for age and HRT for two models: (a) smokers — current or former — compared with never smokers and (b) active smokers compared with no present smokers — never or former Age
HRT status HRT users
HRT never users
n
RR
CI
n
RR
CI
(a) 40–44 45–49 50–54 55–60
253 69 67 57 60
2.4 1.8 1.6 2.7 4.1
1.3–4.4*** 0.5–5.9 0.4–6.2 0.7–10.2 1.4–12.1
262 61 63 59 79
2.1 1.5 8.1 1.2 2.0
1.2 – 3.6*** 0.6 – 3.9 1.7 – 37.8 0.4 – 3.4 0.7 – 5.9
(b) 40–44 45–49 50–54 55–60
107 24 25 29 29
3.2 4.1 3.5 1.9 1.9
1.2–4.8*** 1.0–15.7** 1.2–10.5*** 1.0–3.7* 1.2–3.0*
108 25 26 25 32
2.8 3.0 4.7 1.5 1.6
1.1 – 3.9*** 1.0 – 8.8** 1.7 – 12.7*** 0.9 – 2.5 1.1 – 2.2*
RR, relative risk; CI: 95% confidence interval. * PB0.05. ** PB0.01. *** PB0.005.
or severe WS are shown in Fig. 3. As expected the percentage of subjects who presented moderate – severe WS increased with age in all groups (Figs. 1 – 3). However, a higher proportion of current smokers had moderate to severe WS than life-long non-smokers. Across all age groups, the proportion of former smokers who had moderate – severe WS was higher than life-long non-smokers but lower than for current smokers. HRT did not significantly modify the facial wrinkle score in
current smokers and when comparing HRT users with non-users the only difference observed was among the life-long non-smoker subjects (Table 3). The RR for moderate–severe wrinkling for current smokers adjusted for age was 1.48 (CI: 1.21– 1.80; PB 0.0001) in the HRT non-user group and 1.44 (CI: 1.21–1.72; P B 0.00001) in the HRT user group. The RR of moderate–severe wrinkling for former smokers was 1.09 (CI: 0.95–1.25) in the
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
82
Table 5 Multiple linear regression for wrinkle score (dependent variable) adjusted for packs/year. The effect of smoking on wrinkles’ score was independent of age and HRT Variable
B
SE for B
Beta
T
P-value
Age HRT SM PY (Constant)
1.150746 −4.373457 −2.333396 0.994492 −31.746534
0.093797 1.099150 1.066535 0.083972 4.789901
0.361805 −0.116400 −0.094663 0.514296
12.268 −3.979 −2.188 11.843 −6.628
0.0000 0.0001 0.0290 0.0000 0.0000
Multiple R: 0.61614; R squared: 0.37963; F: 110.91; PB0.0000. SE, standard error; HRT, Hormone Replacement Therapy; SM, Smoking Status; PY: packs/year.
HRT non-user group and 1.08 (CI: 0.97 – 1.21) in the HRT user group. The RR of moderate – severe wrinkling for smokers including both former and current smokers was 1.72 (CI: 1.14 – 2.59; P B0.005) in the HRT non-user group and 2.14 (CI: 1.26 – 3.63; PB 0.002) in the HRT user group (Table 4). Finally, the effect of smoking on wrinkle score was found to be independent of age and HRT (Table 5). Pack-years was positively correlated to facial wrinkles (Fig. 4). Postmenopausal women with pack-year scores over 10 were more prone to have a WS\ 30 (moderate and severe). In analyses adjusted for age, the RR associated with more than 20 pack-years of exposure compared with life-long non-smokers was 1.48 (CI: 1.13 – 1.93; P B0.005) among the HRT users and 1.71 (CI: 1.25 –2.34; PB0.00002) among the HRT nonusers. The proportion of life-long non-smokers on HRT who had a moderate to severe facial WS compared with life-long non-smokers who were HRT non-users was clearly lower (Fig. 3). This difference was more apparent in women over 50 years of age. HRT did not, in general, modify the facial wrinkle score in current smokers, only the group of subjects who smoke or smoked less than 10 pack-years illustrated a protective effect of HRT on WS (Fig. 4). Other variables that may affect facial WS were also studied (Table 6). Sun exposure of 4 h/day or more was associated with higher values of WS (r: 0.579, P B0.01) and with an increased risk of wrinkling. Alcohol consumption ]40 g/day was
also associated with higher values of WS (r: 0.371, PB 0.05). Finally, BMI was inversely related to facial wrinkling (r: − 0.432, PB 0.05).
4. Discussion Great interest has been aroused in therapies that might delay the process of aging. Human aging is a dynamic, unchangeable phenomenon which affects all systems in the body and is commonly perceived as an intrinsic process within a community affecting all of its members. Among the most characteristic aging changes are those undergone by the skin. Skin aging results from an association of chronologic, hormonal (i.e. menopause) and environmental (i.e. tobacco, sun exposure, alcohol, etc.) factors. Classically, chronologic aging includes those cutaneous changes that occur in areas not exposed to the sun, such as the buttocks, and are observed in both men and women. The regressive changes which the skin undergoes with age, such as loss of elasticity, epidermal thickness, and elastic degeneration, may account for the appearance of clinical examples, which include soft tissue sagging and wrinkles. It has been demonstrated that skin collagen is affected by hypoestrogenism [1–3] and that the amount of skin collagen declines in the years following the menopause [1,4–7]. Skin thickness, which was also considered in some of these studies, decreased at a similar rate after the menopause [4–6]. Moreover, the decrease in the amount of skin collagen correlates well with the
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
83
Fig. 4. Relationship between Wrinkle Score (WS) and packs of cigarettes/year according to hormone replacement therapy (HRT) for users (closed circles; dotted line), never users (open circles; dotted line) and total population (solid line). Data from 730 women are shown. Note the significant increase of the WS after 10 packs/year consumption.
reduction in skin thickness observed in the years following castration [8]. Additionally, one of the most significant characteristics of skin biomechanics related to collagen is tensile strength, and several studies have detected a decrease in this parameter with aging [13,14]. Based on these findings, efforts have been made to slow down the degenerative changes in aging skin with oestrogen therapy. As far back as 1946, Goldzieher [15] observed that oestrogens had a regenerative effect on the senile epidermis. This early finding is supported by recent studies which have demonstrated
that oestrogen replacement therapy may increase skin thickness [6,8,16] and skin collagen content [1,4,5]. Even local oestriol may improve the structure of elastic fibres [17]. It is well known that aging skin is more frequently apparent in smokers. Here, we found that current smokers have a RR for facial wrinkling three times that of life-long non-smokers and almost twice that of former smokers. Our results are similar to those reported elsewhere [10,18,19]. This study sought to determine whether HRT neutralized the effects of smoking on the skin.
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
84
Table 6 Relative risk (RR) for moderate or severe facial wrinkling (WS\30) adjusted for age by chosen risk factors HRT status HRT users Variable Agea 40–44 45–49 50–54 55–60 Smoking status Never smoker Former smoker Current smoker Pack/years Never smoker B5 5–10 10–20 \20 Sun exposure B2 h/day 2–4 h/day \4 h/day BMI Lower third Middle third Upper third Alcohol consumption Non-drinker B20 g/day 20–40 g/day \40 g/day
HRT never users RR
CI
n
RR
CI
86 87 89 98
1.0 1.03 1.28 1.53
— 0.93–1.15 1.10–1.49* 1.28–1.83***
85 84 89 112
1.0 2.2 3.1 4.2
— 1.6 – 4.5 2.5 – 6.0** 2.2 – 7.1***
107 150 103
1.0 1.08 1.44
— 0.97–1.21 1.21–1.72**
108 156 106
1.0 1.09 1.48
— 0.95 – 1.25 1.21 – 1.80**
107 25 64 88 76
1.0 0.94 1.03 1.23 1.53
— 0.82–1.08 0.91–1.17 1.05–1.44* 1.24–1.89**
108 36 46 91 89
1.0 0.89 1.14 1.27 1.44
— 0.77 – 1.04 0.92 – 1.42 1.06 – 1.53* 1.17 – 1.79**
72 125 166
1.0 1.36 2.09
— 0.95–2.94* 1.41–3.28**
91 170 106
1.0 1.72 2.26
— 0.87 – 3.19* 1.70 – 4.24**
129 131 133
1.0 0.78 0.64
— 0.45–1.30++ 0.47–1.26+
102 135 130
1.0 0.81 0.69
— 0.50 – 1.59++ 0.47 – 1.18+
35 189 111 28
1.0 0.90 1.15 1.27
— 0.69–1.87 0.85–2.36 1.10–2.90++
45 206 94 22
1.0 0.87 1.20 1.35
— 0.67 – 1.75 0.77 – 1.99 1.12 – 2.97++
n
CI: 95% confidence interval. Not adjusted. ++ PB0.05. + PB0.01. * PB0.001. ** PB0.0001. *** PB0.00000. a
Although this study included a large number of subjects, each of which were evaluated by a single observer, we cannot rule out altogether the possibility of a bias in the scoring for smokers. However, the number of subjects was sufficient to evaluate the effect of smoking and HRT on the one hand with facial wrinkling by age groups on the other, to consider current and former smokers separately and to control for several variables
such as age, pack-years, sun exposure, alcohol consumption and body mass. Our study like others conducted in the same field [10,18,19], still presents certain weaknesses, including the lack of biochemical verification of smoking status, alcohol consumption or HRT use; the fact that several potential confounding factors e.g. the use of sunblocks, sunglasses or other protective products, were not recorded; the possibility of bias in sun
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
exposure across groups; the fact that the wrinkle score was an estimate based only on visual assessment; and finally, the possibility that smoking, use of HRT, and severity of facial wrinkling may be inter-related through other unstudied variables. However, since data were recorded individually in a confidential interview we do not expect the reports of cigarette or alcohol use by subjects to be incorrect. The effect of potential confounding factors, such as skin protective products, might have affected our results, but only if they were associated with smoking status, HRT and wrinkling. Clearly, ultrasonography for measuring skin thickness or profilometry for cutaneous microrelief, as described by Creidi et al. [9], would have been more objective than visual assessment; however, the size of our population meant that the implementation of the two procedures would have been highly complicated. Among smokers (current and former) the RR for wrinkling appeared to increase after 10 packyears. This results, shown in Fig. 4, is in agreement with data from Ernster et al. who found an increased risk for wrinkles in women after the same number of pack-years [10]. We can only hypothesize as to the mechanisms that might be involved in skin aging and facial wrinkling and to their relationship with the use of HRT and the habit of smoking. It has been reported that the ground substance composed mainly of glycosaminoglycans and glycoproteins diminishes in aged skin [20]. Hyaluronic acid is an important constituent and is responsible for the normal turgor of the dermis because of its extraordinary water-holding capacity. The ground substance also contributes to the rheological properties of the skin and may be affected by the components of tobacco. Many of the acute vascular effects of tobacco have been attributed to nicotine [21]; however, tobacco smoke has a great variety of components that may influence directly and indirectly the ground substance. This include tarry and phenolic products, volatile acids and gaseous particles that may account for a chronic hypoxemia of the skin. Moreover, cigarette smoke causes a decrease in the capillary blood flow in the skin [21,22] which may be a causative factor in wrinkling as it affects hyaluronic acid and gly-
85
cosaminoglycan synthesis. To corroborate this hypothesis, further studies are needed. Efforts to reverse the effects of aging skin have been attempted since ancient times, and because of its huge psychological impact, there is much demand for effective treatment. Many well-constructed trials confirm the clinical efficacy of modern therapies with oestrogens for improving the features of aged skin [1,4–6,8,9,15–17]. The effect of estrogens can be seen by increasing glycosaminoglycans and hyaluronic acid concentration, and thereby increasing the hydration of collagenous tissue of the dermis. Despite many claims, there have been no adequate trials documenting the efficacy of fish cartilage polysaccharide extracts or alpha-hydroxy acids, only topical tretionin have been reported to reverse fine wrinkling and mild to moderate hyperpigmentation [23,24]. Although tretionin compounds may be helpful, they also present inconveniences such as skin intolerance, discomfort and a considerable expense [23]. Careful patient selection with consideration of alternative treatments such as chemical peeling [25], dermabrasion, implants [26], and surgery is important in the successful management of skin aging. Finally, our results indicate an accelerated skin aging in current smokers and an improvement in WS only in the life-long non-smoker group, as a result of HRT. Thus, rather than promoting HRT as the solution, smokers need to be made aware of the increased risks of skin aging as a result of their habit. References [1] Castelo-Branco C, Dura´n M, Gonza´lez-Merlo J. Skin collagen changes related to age and hormone replacement therapy. Maturitas 1992;15:113 – 9. [2] Brincat M, Kabalan S, Studd JWW, Moniz CF, de Trafford J, Montgomery J. A study of the decrease of skin collagen content, skin thickness and bone mass in the postmenopausal woman. Obstet Gynecol 1987;70:840 – 5. [3] Brincat M, Moniz CF, Studd JWW, Darby AJ, Magos A, Cooper D. Sex hormones and skin collagen content in postmenopausal women. Br Med J 1983;287:1337 – 8. [4] Brincat M, Versi E, Moniz CJ, Magos A, de Trafford J, Studd JWW. Skin collagen changes in postmenopausal women receiving different regimens of estrogen therapy. Obstet Gynecol 1987;70:123 – 7.
C. Castelo-Branco et al. / Maturitas 29 (1998) 75–86
86
[5] Brincat M, Moniz CF, Kabalan S, Versi E, O’Dowd T, Magos AL, Montgomery J, Studd JWW. Decline in skin collagen content and metacarpial index after the menopause and its prevention with sex hormone replacement. Br J Obstet Gynecol 1987;94(2):126–9. [6] Brincat M, Moniz CF, Studd JWW, Darby AJ, Magos A, Emburey G, Versi E. Long-term effects of the menopause and sex hormones on skin thickness. Br J Obstet Gynecol 1985;92:256 – 9. [7] Castelo-Branco C, Pons F, Gratacos E, Fortuny A, Vanrell JA, Gonza´lez-Merlo J. Relationship between skin collagen and bone changes during aging. Maturitas 1994;18:199 – 206. [8] Punnonen R. Effect of castration and peroral oestrogen therapy on the skin. Acta Obstet Gynecol Scand 1977;21(Suppl):1– 44. [9] Creidi P, Faivre B, Agache P, Richard E, Haudiquet V, Sauvanet JP. Effect of a conjugated oestrogen cream on ageing facial skin. A comparative study with a placebo cream. Maturitas 1994;19:211–23. [10] Ernster VL, Grady D, Miike R, Black D, Selby J, Kerlikowske K. Facial wrinkling in men and women by smoking status. Am J Public Health 1995;85:78–82. [11] Daniell HW. Smoker’s wrinkles. Ann Intern Med 1971;75:873 – 80. [12] Pardell H, Cobo E, Canela J. Manual de Bioestad’stica. Barcelona: Masson Editorial, 1986:183–185. [13] Vogel HG. Influence of maturation and age on mechanical and biomechanical parameters of connective tissue of various organs in the rat. Connect Tissue Res 1978;6:161 – 6. [14] Vogel HG. Attempts to compare ‘in vivo’ and ‘in vitro’ measurements of mechanical properties in rat skin. Bioeng Skin 1981;3:39 –46. [15] Goldzieher MA. The effects of oestrogens on the senile skin. J Gerontol 1946;1:196–201.
.
[16] Punnonnen R, Vilska S, Rauramo L. Skinfold thickness and long-term postmenopausal hormone therapy. Maturitas 1984;5:259 – 62. [17] Punnonnen R, Vaajalahti P, Teisala K. Local oestriol treatment improves the structure of elastic fibers in the skin of postmenopausal women. Ann Chirurg Gynaecol 1987;76(Suppl 202):39 – 41. [18] Kadunce DP, Burr R, Gress R, et al. Cigarette smoking: risk factor for premature facial wrinkling. Ann Intern Med 1991;114:840 – 4. [19] Model D. Smokers’ faces; who are the smokers?. Br Med J 1985;291:1760– 2. [20] Fleishmajer R, Perlish JJ, Bashey RI. Aging of human dermis. In: Pobert CL, editor. Frontiers of Matrix Biology. Basel: Karger, 1973:90 – 106. [21] Reus WF, Robson MC, Zachary L, Heggers JP. Acute effects of tobacco smoking on blood flow in the cutaneous micro-circulation. Br J Plast Surg 1984;37:213 – 5. [22] Richardson D. Effects of tobacco smoke inhalation on capillary blood flow in human skin. Arch Environ Health 1987;42:19 – 25. [23] Noble S, Wagstaff AJ. Tretinoin. A review of its pharmacological properties and clinical efficacy in the topical treatment of photodamaged skin. Drugs Aging 1995;6(6):479 – 96. [24] Helander SD. Treatment of photoaged skin. Efficacy, tolerability and costs of available agents. Drugs Aging 1996;8(1):12 – 6. [25] Humphreys TR, Werth V, Dzubow L, Kligman A. Treatment of photodamaged skin with trichloroacetic acid and topical tretinoin. J Am Acad Dermatol 1996;34:638 – 44. [26] Lemperle G, Hazan-Gauthier N, Lemperle M. PMMA microspheres (Artecoll) for skin and soft-tissue augmentation. Part II: Clinical investigations. Plast Reconstr Surg 1995;96:627 – 34.
.