Structural changes in dermal collagen and oxidative stress levels in the skin of Japanese overweight males

It has been reported that obese people have poorly organized dermal collagen structure because of the degradation of collagen fibers, which is caused by an increase in oxidative stress levels associated with the hypertrophy of subcutaneous adipose cells. However, it is unclear whether an increase in oxidative stress levels caused by the accumulation of subcutaneous adipose tissue and a change in the dermal structure also occur in overweight and obese Japanese people. The objectives of this study are to identify structural changes that occur in the dermis and to measure the levels of oxidative stress in Japanese overweight males.


Synopsis
OBJECTIVE: It has been reported that obese people have poorly organized dermal collagen structure because of the degradation of collagen fibers, which is caused by an increase in oxidative stress levels associated with the hypertrophy of subcutaneous adipose cells. However, it is unclear whether an increase in oxidative stress levels caused by the accumulation of subcutaneous adipose tissue and a change in the dermal structure also occur in overweight and obese Japanese people. The objectives of this study are to identify structural changes that occur in the dermis and to measure the levels of oxidative stress in Japanese overweight males. METHODS: The overweight group included 43 Japanese male volunteers aged between 25 and 64 years and with a body mass index (BMI) of ≥25 and <30. The control group included 47 male volunteers aged between 22 and 64 years and with BMI of <25. The 20-MHz Dermascan C â ultrasound scanner with software for image analyses was used. Echogenicity of the upper and lower dermis was measured. The mRNA expression level of heme oxygenase-1 (HMOX1) in hair follicles was quantitatively analyzed by real-time reverse transcription polymerase chain reaction (RT-PCR) and was used as a marker of oxidative stress. Ultrasonographic imaging and collection of hair follicles were performed at the same site on the thigh, abdomen, and upper arm. RESULTS: The HMOX1 mRNA expression level in the abdomen and thigh was significantly lower in the overweight group than in the control group. Moreover, the echogenicity of the upper dermis of the abdomen and the lower dermis of the abdomen and thigh was significantly lower in the overweight group than in the control group. CONCLUSION: We detected an increase in oxidative stress levels and a decrease in the density of dermal collagen at the same site on the thigh, abdomen, and upper arm of Japanese overweight males. These findings suggest the fragility of the dermis of Japanese overweight males, which might have been caused by the accumulation of subcutaneous adipose tissue.
R esum e OBJECTIF: Il a et e rapport e que les personnes ob eses poss edent une structure mal organis ee des fibres de collag ene en raison de la d egradation du collag ene dermique qui est provoqu ee par une augmentation du stress oxydatif associ e a l'hypertrophie du tissu adipeux sous-cutan e. Cependant, nous ne savons pas si une augmentation du stress oxydatif caus ee par l'accumulation de tissu adipeux sous-cutan e et par une modification de la structure cutan ee se produit egalement chez les personnes japonaises en surpoids et ob eses. Les

Introduction
It has been reported that obese people have poorly organized dermal collagen structure because of the degradation of collagen fibres, and these dermal changes were irreversible. Furthermore, these individuals possess a degraded and diffused collagen matrix compared with normal, healthy controls [1][2][3]. It has also been shown that this dermal condition does not change even after weight loss by bariatric surgery [1]. Moreover, obese individuals have a higher risk of severe pressure ulcer and wound infection, as well as require a longer time for post-operative wound healing [4,5] and longer hospitalization. These findings thus suggest that a structural change in the dermis of obese individuals might cause a decline in their quality of life. Therefore, it is important to determine the changes in the dermal structure that occur in obese individuals to design an appropriate approach for skincare in these individuals. However, structural changes occurring in the skin of obese individuals and the mechanisms underlying these changes were not been established.
Using animal models, we describe that an increase in oxidative stress levels due to the hypertrophy of subcutaneous adipose cells, which produce collagenolytic enzymes, leads to structural changes in the dermis. We also observed a thickening of the subcutaneous adipose tissue as well as a decrease in the density and convolution of the collagen fibres in the dermis of obese diabetic mice compared with control mice [6][7][8]. Moreover, we observed an increase in the mRNA expression level of heme oxygenase 1 (HMOX1), a marker of oxidative stress, and matrix metalloproteinase (MMP), an enzyme that degrades type I collagen, as well as a decrease in the skin tensile strength of obese diabetic mice [6]. In antioxidant experiments to determine whether the change in collagen fibres and skin fragility was caused by oxidative stress, oral administration of an antioxidant markedly decreased the expression levels of HMOX1 and MMP and improved the skin tensile strength and the structure of collagen fibres in obese diabetic mice [6]. These findings suggest that skin fragility in obese diabetic mice is associated with dermal collagen damage and weakened tensile strength and that oxidative stress and MMP overexpression in the subcutaneous adipose tissue may, at least in part, affect dermal fragility via a paracrine pathway.
Investigations involving the skin of obese individuals with a body mass index (BMI) of ≥30 and obese animal models have been conducted. However, studies relating to overweight individuals with 25 ≤ BMI < 30 are limited, despite the possibility that accumulation of subcutaneous adipose tissue may lead to skin changes in this particular subpopulation. 'Obesity in Japanese' is defined as BMI of ≥25 kg m À2 because Japanese are at high risk of hypertension and diabetes even at low BMI, and most of the cases of obesity in Japan correspond to overweight people with 25 ≤ BMI < 30 [9]. For Asians, insulin resistance [10][11][12] and the optimal cut-off points for BMI for discriminating diabetes and hypertension are lower (24-25 kg m À2 ) than those for Caucasians (26-28 kg m À2 ); these may explain the similar prevalence rate of diabetes between both races. Similar to obese individuals, overweight people accumulate more subcutaneous adipose tissue than individuals of normal weight. Based on the association between visceral fat accumulation and the risk of diabetes and cardiovascular diseases [13][14][15], we thus aimed to determine the relationship between the deposition of subcutaneous adipose tissue and structural changes in the skin of overweight Japanese individuals. There is currently no specialized skincare regimen for overweight individuals in Japan as well as the rest of the world because skin problems are rarely manifested in this particular subpopulation. However, microscopic signs of skin fragility have been observed in overweight people as well as in animal models, which indicate the need for skincare to decrease the incidence of severe pressure ulcers and to improve the post-operative wound healing period of this particular subpopulation [4,5].
We previously reported that Japanese overweight males have higher oxidative stress levels in the skin and a lower high echo spot in the lower dermis, as detected by ultrasonographic imaging of the thigh [16]. We performed a non-invasive assessment of human skin by visualizing the structure of the human dermis using ultrasonographic imaging and by analysing oxidative stress based on the mRNA expression levels in cells from hair follicles. In the Japanese overweight males, the echogenicity of the lower dermis decreased, whereas the thickness of the dermis increased [16]. It is possible that the decrease in dermal echogenicity is reflective of the decrease in the density of dermal collagen fibres.
Although an increase in oxidative stress levels in the skin may cause structural changes in the dermis, it is not possible to measure oxidative stress using a single region of the body because several factors influence oxidative stress [17]. In addition, in our previous study, we did not investigate the ultrasonographic image and oxidative stress levels in the skin of the same site; extraction of hair follicles was performed on the front of the thigh, and the assessment of the dermal structure by ultrasonographic imaging was performed on the back of the thigh. We need an assessment scheme to clearly show whether the increase in oxidative stress levels in the skin and structural changes in the dermis have actually occurred on the same part.
This study aimed to determine whether skin fragility also occurs in the same body site and whether skin fragility occurs in body parts other than the thigh. Therefore, the objectives of this study were to identify structural changes that occur in the dermis and to measure oxidative stress levels in Japanese overweight males using ultrasonographic imaging of hair follicles collected from the thigh and other body parts.

Study design and setting
This cross-sectional study was conducted from April to November 2013. Data collection was performed at the Kanazawa University campus. All data analyses were performed at the Kanazawa University campus.

Subjects
The study population included healthy volunteer males with 25 ≤ BMI < 30 kg m À2 ; these individuals consisted of staff members or students of the university, employees of a non-financial company and neighbouring residents of Kanazawa University. We also sought healthy volunteer males with BMI of <25 kg m À2 for the control group. The subjects belonged the age range of 20-64 years. Subjects with a sharp pain or rubor in the body part that would be used for the measurements of oxidative stress levels and dermal changes and those with a systemic skin disorder, dry skin, oedema, alcoholism, liver cirrhosis, hepatic insufficiency and renal insufficiency were excluded.

Measurements and instruments
Demographics and characteristics of the study population Information on age, sex, anamnesis, oral medication intake and drinking and smoking habits were collected by oral consultation of subjects. We also measured their height and body weight. We calculated BMI (kg m À2 ) and body fat ratio (%) to determine the degree of obesity of each study participant. Bioelectrical impedance analysis was performed using a multifrequency body composition metre (MC-190; Tanita, Tokyo, Japan) to measure body weight and body fat ratio. We also measured the waist circumference diameter, which is an index of visceral fat accumulation, using a measuring tape. We also interviewed the subjects whether they had diabetes, hyperglycaemia, hypertension and hyperlipemia, as well as any history of smoking.

Measurement of research parameters
We examined the upper arm, abdomen and thigh for any accumulation of subcutaneous adipose tissue, as well as the existence of body hair. For the upper arm, we examined the exterior of the left upper arm, whereas the study participant was in a prone position. A photograph was taken of this body region, and a hair follicle was extracted from the mid-point of a straight line that connected the left elbow to the left acromion. For the abdomen, we examined the navel circumference, whereas the study participant was in a dorsal recumbent position. A photograph of the body region was taken, and a hair follicle was extracted from a region 2 cm to the right of the navel. For the thigh, we examined the rear side of left thigh, wheras the study participant was in a prone position. A photograph of the body region was taken, and a hair follicle was collected from that particular body part at maximum thigh girth, at a height equivalent to one half of a straight line that connected the centre of the left knee socket to the left of the trochanter major.

Structure of dermis
The 20-MHz Dermascan C â ultrasound scanner (CortexTechnology, Hadsund, Denmark) equipped with software for image analyses was used for the assessment of dermal structure [18][19][20][21]. The 20-MHz ultrasound transducer allows a 60 9 15-lm resolution and a 13-mm penetration depth. The ultrasound velocity for skin was set at 1580 m s À1 [19]. The probe for the longitudinal scan was placed on every site. A total of three photographs were acquired per site in the B-mode. All photographs were taken using fixed values for the gain profile level 3 and the gain profile level 10.
Echo-free structures were displayed as black regions in the ultrasound image. Connective tissue structures appeared in green, red or yellow. Because the ultrasound reflection intensity is related to the relative density of the targeted tissue, it also provides information on the arrangement of the collagen and elastic fibres. The colour scale indicates the intensity of ultrasound reflection, wherein white indicates the highest reflection, and black indicates the lowest reflection [21].

Structure of subcutaneous adipose tissue
The 18-MHz Mylab TM five â ultrasound scanner (Esaote, Genoa, Italy) was used for the assessment of subcutaneous adipose tissue. The probe for the longitudinal scan was placed on every site. The gain was adjusted for each individual, and a photograph was taken. A total of three photographs were acquired per site.

Oxidative stress
After cutting the hair short with surgical clipper (3M TM , St. Paul, MN, U.S.A.), three hair follicles that included the outer root sheath (ORS) were collected using tweezers from each body part of the study subjects. The collected samples were stored in 1 mL of RNAlater (R0901-100ML; Sigma-Aldrich, St. Louis, MO, U.S.A.) at 4°C until RNA extraction. Total RNA was extracted from ORS tissue using the RNeasy Plus Micro Kit (Qiagen, Venlo, the Netherlands), following the manufacturer's instruction. The extracted RNA samples were maintained at À80°C until reverse transcription. cDNA was synthesized from each RNA sample using the QuantiTect Reverse Transcription Kit (Qiagen), following the manufacturer's instruction.
We used the relative expression level of HMOX1 mRNA [22] as a marker of oxidative stress in the local skin of the thigh, abdomen and upper arm. The relative expression level of Hmox1 The amplification conditions were as follows: 95°C for 30 s and 60°C for 1 min was repeated for 50 cycles after preheating at 95°C for 10 min. The HMOX1 mRNA expression level was quantified by the comparative Ct method. PCR was performed twice per sample, and computation for the average Ct values as a final Ct value was performed twice.

Image analysis
For the evaluation of the density of dermal collagen, we used the built-in image analysis software of Dermascan C â (CortexTechnology). We measured the echogenicity of the upper and lower dermis of the same area of 0.754 (0.58 mm 9 1.3 mm) mm 2 to evaluate the density of dermal collagen. Because segmentation analysis of the clinical images could be immediately and easily performed using this tool, we identified the region-of-interests (ROI) at the bottom of the epidermis and at the top of the subcutaneous adipose tissue by superposing a 'Shape1 rectangle', from which the 'Total intensity in%' was measured. Total intensity in% means the average value of the echogenicity (0-255 range of pixel intensity) of ROI by percentage (Fig. 1). Earlier investigations have shown that the echogenic density of ultrasound is high, based on the quantity of collagen and density of dermal collagen [21,23,24]. The evaluation of echogenicity using Dermascan C â has also been used in previous studies [19,20,23,[25][26][27]. The echogenicity of the upper and lower dermis has also been used for the evaluation of papillary and reticular layer of dermis in a previous study [28]. For the evaluation of the thickness of the dermis and subcutaneous adipose tissue, we used the built-in measurement tool of Mylab five â (Esaote). Because this image captured by this instrument clearly identifies the subcutaneous adipose tissue, we judged that this instrument could accurately measure the thickness of the dermis, which is a specific layer of the subcutaneous adipose tissue. To calculate the thickness of the subcutaneous adipose tissue and dermis, we computed for the average thickness of three areas per image.

Statistical analysis
All analyses were performed using SPSS (Chicago, IL, U.S.A.) for Windows ver. 21.0. Descriptive data were expressed as the mean AE standard deviation for continuous variables and n for categorical variables. The subjects were classified into two groups according to BMI < 25 and 25 ≤ BMI < 30, representing the control and overweight groups, respectively. These two groups were compared using independent t-test, Mann-Whitney U-test, chisquare test and Fisher's exact test. In this study, a p value of <0.05 was considered significant.

Ethical considerations
This study was approved by the Ethics Committee of the Kanazawa University School of Medicine. Information obtained from the subjects was not used for any other purpose than the study objective. The privacy of the subjects was protected, and the subjects were not identified. Participation in the study was voluntary, and no disadvantage was incurred by refusing to participate.

Demographics and characteristics of study population
A total of 90 male volunteers were recruited (Table I). The average age of the 43 overweight subjects with 25 ≤ BMI < 30 was 41.6 AE 11.8(age  and that of the 47 control subjects with BMI < 25 was 38.3 AE 13.8(age . In terms of the basic characteristics of the subjects, the values for body weight, body fat ratio, waist circumference diameter and the number of individuals with hyperlipemia and metabolic syndrome were significantly higher in the overweight group than in the control group.

Ultrasonographic imaging of dermis per body site
Representative ultrasonographic images are shown in Figs 2 and 3. The echogenicity of the upper and lower dermis was significantly lower in the overweight group than in the control group (Fig. 2). Moreover, the dermis in the overweight group was relatively thicker than that in the control group (Fig. 3).

Evaluation of the skin parameters of the study participants
The thickness of the dermis and subcutaneous adipose tissue in the three body regions was significantly higher in the overweight group than in the control group. Moreover, the HMOX1 mRNA expression level in the abdomen and thigh was significantly higher in the overweight group than in the control group. Furthermore, the echogenicity of the upper dermis of the abdomen and the lower dermis of the abdomen and thigh was significantly lower in the overweight group than in the control group. The HMOX1 mRNA expression level was higher and the echogenicity of the lower and upper dermis was lower in the upper arm of the overweight group than in the control group, although these findings were not significant (Table II).
We were unable to collect the hair follicle samples from some of the subjects because of reduced hair growth on their skin. The efficiency of hair follicle collection from all subjects and the efficiency of quantification of the HMOX1 mRNA expression level, which  accounts for all the hair follicle samples, thus varied among body sites. The number of the results for the HMOX1 mRNA expression level thus also varied.
The efficiency of hair follicle collection for the control group was 89.1%, 97.9% and 36.4% for the abdomen, thigh and upper arm, respectively. The efficiency for the overweight group was 90.7%, 100.0% and 37.2% for the abdomen, thigh and the upper arm, respectively. The efficiency of quantification of the HMOX1 mRNA expression level in the hair follicle samples extracted from the control group was 75.6%, 87.0% and 50.0% for the abdomen, thigh and upper arm, respectively. The efficiency of quantification for the overweight group was 82.1%, 88.4% and 60.0% for the abdomen, thigh and upper arm, respectively. For the abdomen and thigh, the HMOX1 mRNA expression level was significantly higher in the overweight group than in the control group.

Discussion
In the present study, we investigated the structure of the dermis of the thigh, abdomen and upper arm using ultrasonographic imaging, as well as examined oxidative stress levels in the skin using hair follicles from the same body regions of Japanese overweight males. This present study has two originalities. We detected an increase in the oxidative stress levels in the skin from the three body regions. We also observed a decrease in dermal echogenicity using ultrasonographic imaging at the three sites, which was indicative of the oxidative stress. The thickening of the subcutaneous adipose tissue might have increased the oxidative stress levels. Although we cannot deny the possibility that the mechanical stress induced by clothes on the thigh and the stress caused by ultraviolet rays on the upper arm might have also influenced the oxidative stress levels in the skin, the observed increase in oxidative stress levels in the skin of the abdomen may be a direct response to the accumulation of subcutaneous adipose tissue because this body region is less likely subjected to other external stressors. However, the extraction possibility rate and the quantification possibility rate of hair follicle samples from the upper arm were clearly low in this study, suggesting that this was not a suitable region to measure oxidative stress using hair follicles in Japanese overweight males.
The decrease in dermal echogenicity is suggestive of a lower density of dermal collagen fibres. In previous studies using skin samples from humans or dogs, the dermal collagen content of the skin showed a positive correlation with ultrasound speed, which is an ultrasonic attenuation coefficient of skin tensile strength [29,30]. The lower dermal echogenicity values obtained for overweight individuals were in agreement with the findings of a previous study that employed the same ultrasound scanner [16], which further strengthens our hypothesis of a decrease in density and convolution of the collagen fibres in the dermis [6][7][8]. And De Rigal et al. [24] are contrasting the histology of skin and the dermal ultrasonographic image. They described that the density of collagen in reticular dermis is high and this result is in agreement with the high echogenicity of lower dermis in ultrasonographic image. Therefore, we consider this high echogenicity band on the lower dermis reflecting the portion dermal collagen is density as the previous study.
Therefore, we believe that we can quantify the density of dermal collagen in overweight individuals by measuring echogenicity using ultrasonographic imaging. The collagen component around the subcutaneous adipose tissue in the abdomen is generally poor in the thigh [30], thus resulting in variations per body site. A previous study showed that the echographic density of the dermis in some parts of the body might have a negative correlation with the thickness of the dermis [25], which is supported by the results of the present study.
The decrease in dermal echogenicity may be attributable to the increase in oxidative stress due to the accumulation of subcutaneous adipose tissue. Our findings of an increase in the HMOX1 mRNA expression level in the skin and a decrease in the density of dermal collagen were in agreement with the earlier findings involv- We were unable to collect the hair follicle samples from some of the subjects because of reduced hair growth on their skin. The efficiency of hair follicle collection from all subjects and the efficiency of quantification of the HMOX1 mRNA expression level, which accounts for all the hair follicle samples, thus varied among body sites. The number of the results for the HMOX1 mRNA expression level thus also varied. *Independent t-test.
ing obese diabetic mice [8,10]. We have analysed the expression of MMPs in the skin of obese individuals using in vitro and in vivo methods and identified their overexpression using two approaches: through the deposition of fat itself as an indicator of overexpression [31] and by measuring oxidative stress levels accompanying ischaemia and low oxygen levels due to hypertrophy of adipose cells [6,8]. In the present study, we did not evaluate the expression levels of MMPs, and the course of the increase in oxidative stress in the Japanese overweight skin remains unknown. We would like to investigate this area in our future research studies. This study has some limitations, and the results of this study are not necessarily applicable to all overweight individuals. First, there are racial differences. In general, Asians have a thicker dermis than Caucasians [32], and Asian skin is less likely affected by ageing [33]. Second, there are sex differences. Males have a thicker dermis [34], and the female dermis tends to produce cellulite [21,22,35]. Third, the skin hydration is an important factor which may influence echogenicity. And, we did not measure the skin hydration in this present study. It has been reported high-frequency ultrasound effective in presumption of skin hydration. For example, it has been reported dermal echoge-nicity decreased by the dry skin of the face and skin oedema [28,36]. As we did not include the subjects with dry skin or skin oedema in the present study, we think that the influence of echogenicity by the skin hydration is the minimum. Moreover, in our unpublished data, dermal echogenicity decreases with higher BMI in females. In the future, we plan to develop an assessment technology that evaluates oxidative stress levels in the skin of Japanese women from whom hair follicle extraction is generally difficult.

Conclusions
We detected an increase in oxidative stress levels and a decrease in the density of dermal collagen at the same site on the thigh, abdomen and upper arm of Japanese overweight males. These findings suggest the fragility of the dermis of Japanese overweight males, which might have been caused by accumulation of subcutaneous adipose tissue.