Abstract
The objective of this study was to construct a predictive model to identify aged care residents at risk of future skin tears. Extensive data about individual characteristics, skin characteristics, and skin properties were gathered from 173 participants at baseline and at 6 months. A predictive model, developed using multivariable logistic regression, identified five variables that significantly predicted the risk of skin tear at 6 months. These included: a history of skin tears in the previous 12 months (OR 3.82 [1.64‐8.90], P = 0.002), purpura ≤20 mm in size (OR 3.64 [1.42‐9.35], P = 0.007), a history of falls in the previous 3 months (OR 3.37 [1.54‐7.41], P = 0.002), clinical manifestations of elastosis (OR 3.19 [1.38‐7.38], P = 0.007), and male gender (OR 3.08 [1.22‐7.77], P = 0.017). The predictive model yielded an area under the receiver operating characteristic curve of 0.854 with an 81.7% sensitivity and an 81.4% specificity. This predictive model could inform a simple but promising bedside tool for identifying older individuals at risk of skin tears.
Keywords: ageing skin, clinical skin manifestations, cohort study, predictive model, skin tears
1. BACKGROUND
Skin tears comprise a substantial proportion of all wounds found amongst older adults.1, 2, 3, 4, 5, 6 They are commonly defined as “a traumatic wound occurring principally on the extremities of older adults, as a result of friction alone or shearing and friction forces which separate the epidermis from the dermis (partial‐thickness wound) or which separates both the epidermis and the dermis from underlying structures (full‐thickness wound)”.7 Published epidemiological data on the prevalence and incidence of skin tears in aged care facilities indicate that they occur in 41% to 59% of Australian, 14% to 22% of North American, and 4% to 14% of Japanese residents.1, 3, 8, 9, 10, 11 Regardless of the geographical location of published studies, skin tears principally occur on the upper extremities, followed by the lower extremities, of older individuals.2, 8, 12, 13
A recent review of the literature identifying individual attributes or clinical characteristics that predicted the risk of skin tears in older adults found that the vast majority of articles were based on lower levels of scientific evidence including expert opinion, case series, and observational studies.14 Three studies that used advanced statistical modelling to predict the risk of tears were identified.9, 15, 16 The first study by Lewin et al.15 used multivariate regression analyses of data obtained from a non‐matched case–control study that was conducted in a Western Australian tertiary hospital on participants aged over 50 years.15 The predictive model identified five skin characteristics (presence of senile purpura, bruising, haematoma, evidence of healed skin tears, and oedema) and a single individual factor (inability to reposition independently) that significantly predicted the risk of skin tears. A prospective cohort study undertaken at the same tertiary facility by Newall et al.16 tested these variables and found that the model had high sensitivity but low specificity for predicting skin tears.16 Secondary analysis was undertaken using stepwise logistic regression of data from the initial case–control study combined with the prospective cohort study, which was subsequently randomly split into two and generated a revised model from one half of the dataset. The revised model reported that senile purpura, haematoma, previously healed skin tears, advanced age, and the ability to reposition independently were better predictors of skin tears.16 The third study by Sanada et al. in 2015 used multiple logistic analyses of data collected from a 3‐month prospective cohort study conducted in Japan and identified that a history of skin tears and a decreased Braden Scale score were significant predictors of skin tears.9 The analysis was based on a small number (n = 14) of skin tears that developed during the study period.
The review revealed a paucity of clinical studies that predicted the risk of skin tears in older individuals.14 There is clearly a clinical need to better understand associated aetiologies and accurately predict the risk of skin tears amongst elderly individuals. A reliable skin tear predictive model would permit more effective targeting of preventative interventions to reduce the incidents of these injuries. However, accurate prediction of individuals at risk of skin tears needs to be based on a standardised, comprehensive assessment that captures a broad range of quantified factors. The objective of this study was therefore to construct a predictive model to identify older individuals at risk of tears using quantified data of individual characteristics, clinical characteristics and manifestations, and morphological and physiological skin properties. These data were obtained from a prospective cohort study, which was carried out between February 2014 and June 2015 in Western Australia across four aged care facilities to identify variables that were significantly associated with the 6‐month risk of skin tears. The demographic and clinical profiles of these participants have been reported.17
2. MATERIALS AND METHOD
2.1. Study design and setting
This study is the final of a three‐stage research that comprised a preparatory stage, a pilot study, and the major study. The methodology and results of the preparatory stage and pilot study have previously been reported.18 This major study followed the same methodology applied in the pilot study,16 but the research was conducted on a much larger sample of aged care residents over a 6‐month period. Two Western Australian regional and two metropolitan aged care facilities that were operated by a single not‐for‐profit service provider participated in this prospective cohort study.
Ethics approval for all stages of this study was obtained from the Curtin University Research and Development Human Research Ethics Committee (RD‐23‐13) and the Bethanie Group Inc. Governance Committee. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and the Australian Code for the Responsible Conduct of research.19, 20
2.2. Measurements and instruments
Two commercially available, non‐invasive instruments (DermaLab Combo, and Skin‐pH‐meter) were used to assess ageing morphological and physiological skin properties. The DermaLab Combo (Cortex Technology, Hadsund, Denmark), a multi‐purpose device, was used to assess transepidermal water loss (TEWL), hydration, skin thickness, and elasticity. The Skin‐pH‐meter (Courage + Khazaka, Cologne, Germany) device was used to evaluate skin surface pH. A detailed explanation of these instruments and the various skin properties was provided in the pilot study.18
Collaboration with researchers from the Department of Gerontological Nursing and Wound Care Management at the University of Tokyo facilitated the analysis of three transepidermal secreted skin proteins (collagen type IV, matrix metalloproteinase‐2 (MMP‐2), and tumour necrosis factor‐alpha (TNF‐α)). These proteins were collected from one upper and lower extremity using a skin blotting technique devised by researchers at the University of Tokyo.21 The proteins were transepidermally collected using non‐invasive skin blotting with nitrocellulose material that was applied to the skin for 10 minutes.13 Koyano et al.13 had previously implicated these secreted proteins in the risk of skin tears. Subsequent research by the authors failed to identify any significant association with skin tears.10
2.3. Participants
A total of 200 aged care residents participated in this prospective cohort study. The size of the sample was based on the statistical power to estimate the reliability of test–retest measurements.22 An estimated minimum sample size of 160 participants was required to provide 80% statistical power at the 5% level of significance to detect effects from subsequent data analysis and statistical modelling of skin properties as measured by the non‐invasive devices at two points in time.23, 24 An additional 40 participants were recruited for the study to account for potential dropout from death, transfer of participants, or withdrawal from the study. Residents were recruited to this study if they were aged over 65 years and where informed written consent was provided either by themselves or their legal guardian. Residents were excluded if: consent was not obtained, they had a connective tissue disorder, were in pain, were agitated at the time of assessment, had a lower limb amputation, or were receiving palliative services.
Data derived from the 31 residents who participated in the pilot study were pooled with an additional 169 recruited residents to form the major study's baseline measurements.18 Pooling of data was considered appropriate as the design and methodology of the major study replicated that of the pilot study, with only minor refinements. The data were also combined to reduce the need for unnecessary examinations and to minimise any potential discomfort that multiple assessments could potentially pose to participants.
2.4. Data collection
Data were collected on a broad range of individual characteristics, skin characteristics, and biophysical skin properties by a single investigator at two points in time, 6 months apart. The selection of individual and skin characteristics was based on the skin tear literature and factors that were reported to contribute to cutaneous ageing.14, 15, 25
For the purpose of this study, non‐invasive testing of biophysical skin properties were collected from four anatomical test sites: bilateral mid‐dorsal forearms (midpoint between lateral epicondyle and radial styloid process) and the upper quartile of the lateral lower legs. The extremities had previously been identified as being vulnerable to skin tears.8, 12 The upper quartile of the lateral lower extremities was selected following technical difficulties identified in the preliminary study.18
Except for skin blotting, where a single measurement was collected at the initial assessment, three consecutive measurements were taken 10 mm apart at the designated test sites for each skin property (for reliability control), with the combined mean used for calculation. The skin tear occurrence data, which recorded the number and location of skin tears, for all participants were obtained from the service providers’ integrated database 6 months after the initial skin assessment.
2.5. Statistical analysis
The data analysis was undertaken in two parts. The initial analysis determined the intra‐class correlation coefficient (ICC) as estimates of the intra‐rater reliability for the pilot study over a 6‐month period. The ICC results for the pilot study have previously been reported.18 Subsequent analysis was conducted to identify baseline variables that significantly predicted the risk of skin tears at 6 months.
Descriptive statistics for individual characteristics and skin characteristics were calculated as the mean, standard deviation, and median (interquartile range) for continuous variables and frequency (percentage) for categorical variables. Chi‐squared tests were conducted to evaluate the frequency data for categorical variables and the occurrence of skin tears at 6 months.26 Independent sample t‐tests were used to compare continuous data between participants with skin tears and those without skin tears.
Univariable logistic regression was used to identify variables significantly associated individually with the risk of skin tears at 6 months. All variables were considered, in multivariable logistic regression analysis, to develop a parsimonious model with the least number of independent and statistically significant risk variables27 after taking into consideration possible confounders.28 The magnitude of the associations between each independent variable and the outcomes was expressed as an odds ratio (OR) with 95% confidence intervals (CIs) obtained by multivariable logistic regression. The OR measured the strength of association between a variable and the 6‐month risk of skin tears.29
The ROC curve was computed using the sensitivity and specificity of the model in discrimination testing of residents with and without skin tears, and the corresponding area under the curve (AUC) provided a graphic illustration of the performance of the statistical model.30 The AUC was used to interpret the probability that the skin tear model correctly identified participants with and without skin tears. The values for AUC ranged from 0 to 1, with values of <0.5 indicating no discrimination, 0.6 poor discrimination, 0.7 acceptable/good discrimination, 0.8 very good discrimination, and 0.9 and above as excellent discrimination.31 A predictive model ideally needs to demonstrate both high sensitivity and specificity and, consequently, a high AUC.31
3. RESULTS
Data were collected between February 2014 and June 2015 from 200 participants at baseline and 173 of the same residents 6 months later. Of the original 200 participants, 27 (13.5%) were lost to follow up (25 participants were deceased during the 6 months, and 2 participants declined to take part further in the study). The 173 reassessed participants comprised of 123 (71%) females and 50 (29%) males with a combined mean age of 87.6 years (SD ± 6.7). Of these participants, 20 (11.6%) had a skin tear at baseline, with 14 (11.4%) females and 6 (12%) males having this type of wound.
Table 1 presents the baseline‐documented individual and skin characters by the number and relative proportion of participants who sustained a skin tear during the 6‐month period.
Table 1.
Comparison between participants with and without skin tears at 6 months based on personal characteristics and skin characteristics variables, as reported in the literature
| Characteristics | Total | Participants with skin tears (n = 71) | Participants without skin tears (n = 102) | P‐value |
|---|---|---|---|---|
| n (%) | n (%) | |||
| Individual characteristics | ||||
| Age: Mean (SD) | 87.6 (6.7) | 88.6 (6.1) | 86.6 (7.2) | 0.049* |
| Gender | ||||
| Males | 50 | 31 (62.0) | 19 (38.0) | <0.001* |
| Females | 123 | 40 (32.5) | 83 (67.5) | |
| Fitzpatrick skin type | ||||
| Type 1 | 9 | 2 (22.2) | 7 (77.8) | 0.705 |
| Type 2 | 52 | 22 (42.3) | 30 (57.7) | |
| Type 3 | 78 | 33 (42.3) | 45 (57.7) | |
| Type 4 | 34 | 14 (41.2) | 20 (58.8) | |
| Place of birth | ||||
| Oceanian | 119 | 53 (44.5) | 66 (55.5) | 0.577 |
| North‐west European | 38 | 12 (31.6) | 26 (68.4) | |
| Southern/eastern European | 7 | 3 (42.9) | 4 (57.1) | |
| South‐east Asian | 2 | 1 (50.0) | 1 (50.0) | |
| North‐east Asian | 1 | 0 (0.0) | 1 (100) | |
| Southern and central Asian | 1 | 1 (100) | 0 (0.0) | |
| Central Americans | 3 | 1 (33.3) | 2 (66.7) | |
| Sub‐Saharan African | 2 | 0 (0.0) | 2 (100) | |
| Number of years living in Australia: Mean (SD) | 77.2 (20.8) | 80.1 (19.9) | 74.3 (21.7) | 0.077 |
| Body mass index | 25.6 (5.4) | 25.8 (5.8) | 25.4 (5.0) | 0.635 |
| Weight in kilograms | 67.4 (16.1) | 69.6 (17.1) | 65.1 (15.1) | 0.070 |
| Height in centimetres | 162.5 (162.5) | 164.0 (9.6) | 160.9 (8.1) | 0.022 |
| ADL score | 80.2 (17.3) | 80.7 (17.3) | 79.7 (17.2) | 0.718 |
| Braden Scale score | 18.2 (3.5) | 17.8 (3.3) | 18.4 (3.7) | 0.296 |
| Principle work environment | ||||
| Mix indoor/outdoor work | 150 | 52 (34.7) | 98 (65.3) | <0.001 |
| Primarily worked outdoors | 23 | 19 (82.6) | 4 (17.4) | |
| Ability to reposition self | ||||
| No | 91 | 44 (48.4) | 47 (51.6) | 0.039* |
| Yes | 82 | 27 (32.9) | 55 (67.1) | |
| History of smoking | ||||
| Lifelong non‐smoker | 96 | 34 (35.4) | 62 (64.6) | 0.229 |
| Ex‐smoker | 67 | 30 (44.8) | 37 (55.2) | |
| Unknown | 10 | 0 (0.0) | 0 (0.0) | |
| Number of chronic diseases | 3.7 | 3.7 (1.4) | 3.6 (1.6) | 0.832 |
| Nutrition | ||||
| Well nourished | 129 | 49 (38.0) | 80 (62.0) | 0.353 |
| Resident obese | 33 | 17 (51.5) | 16 (48.5) | |
| Underweight and frail | 11 | 5 (45.5) | 6 (54.5) | |
| Contractures | ||||
| No | 167 | 69 (41.3) | 98 (58.7) | 0.696 |
| Yes | 6 | 2 (33.3) | 4 (66.7) | |
| Paralysis | ||||
| No | 166 | 71 (42.8) | 95 (57.2) | 0.024 |
| Yes | 7 | 0 (0.0) | 7 (100) | |
| Dementia | ||||
| No | 77 | 34 (44.2) | 43 (55.8) | 0.456 |
| Yes | 96 | 37 (38.5) | 59 (61.5) | |
| Heart disease | ||||
| No | 100 | 43 (43.0) | 57 (57.0) | 0.540 |
| Yes | 73 | 28 (38.4) | 45 (61.6) | |
| Respiratory disease | ||||
| No | 152 | 58 (38.2) | 94 (61.8) | 0.038* |
| Yes | 21 | 13 (61.9) | 8 (38.1) | |
| Renal disease | ||||
| No | 155 | 63 (40.6) | 92 (59.4) | |
| Yes | 18 | 8 (44.4) | 10 (55.6) | 0.756 |
| Activity level | ||||
| Walks frequently | 49 | 27 (35.5) | 22 (64.5) | 0.096 |
| Walks occasionally | 64 | 33 (51.6) | 31 (48.4) | |
| Bedfast and chairfast | 33 | 11 (33.3) | 22 (66.7) | |
| Agitation | ||||
| None or occasional issue | 72 | 23 (31.9) | 49 (68.1) | 0.040 |
| Moderate to severe issue | 101 | 48 (47.5) | 53 (52.5) | |
| Falls risk category | ||||
| Medium (11‐20) | 83 | 22 (26.5) | 61 (73.5) | <0.001 |
| High (21‐39) | 90 | 49 (54.4) | 41 (45.6) | |
| History of falls at 1 month | ||||
| No | 127 | 46 (36.2) | 81 (63.8) | 0.032* |
| Yes | 46 | 25 (54.3) | 21 (45.7) | |
| History of falls at 3 months | ||||
| No | 100 | 28 (28.0) | 72 (72.0) | <0.001* |
| Yes | 73 | 43 (58.9) | 30 (41.1) | |
| History of falls at 6 months | ||||
| No | 73 | 18 (24.7) | 55 (75.3) | <0.001* |
| Yes | 100 | 53 (53.0) | 47 (47.0) | |
| Total number of medications: Mean (SD) | 6.9 | 6.8 (3.5) | 7.0 (3.4) | 0.678 |
| Corticosteroid medications | ||||
| No | 119 | 47 (39.5) | 72 (60.5) | 0.540 |
| Yes | 54 | 24 (44.4) | 30 (55.6) | |
| Anticoagulants | ||||
| No | 164 | 69 (42.1) | 95 (57.9) | 0.239 |
| Yes | 9 | 2 (22.2) | 7(77.8) | |
| Antiplatelets | ||||
| No | 103 | 39 (37.9) | 64 (62.1) | 0.303 |
| Yes | 70 | 32 (45.7) | 38 (54.3) | |
| Oral corticosteroids | ||||
| No | 163 | 66 (40.5) | 97 (59.5) | 0.553 |
| Yes | 10 | 5 (50.0) | 5 (50.0) | |
| Topical corticosteroids | ||||
| No | 135 | 54 (40.0) | 81 (60.0) | 0.600 |
| Yes | 38 | 17 (44.7) | 21 (55.3) | |
| Inhalation corticosteroids | ||||
| No | 160 | 65 (40.6) | 95 (59.4) | 0.697 |
| Yes | 13 | 6 (46.2) | 7 (53.8) | |
| Sedative | ||||
| No | 144 | 61 (42.4) | 83 (57.6) | 0.431 |
| Yes | 29 | 10 (34.5) | 19 (65.5) | |
| Antihypertensives | ||||
| No | 82 | 32 (39.0) | 50 (61.0) | 0.609 |
| Yes | 91 | 39 (42.9) | 52 (57.1) | |
| Antiplatelets | ||||
| No | 103 | 39 (37.9) | 64 (62.1) | 0.303 |
| Yes | 70 | 32 (45.7) | 38 (54.3) | |
| Moisturiser used | ||||
| No | 57 | 25 (43.9) | 32 (56.1) | 0.597 |
| Yes | 116 | 46 (39.7) | 70 (60.3) | |
| History skin tears at 3 months | ||||
| No | 132 | 45 (34.1) | 87 (65.9) | 0.001* |
| Yes | 41 | 26 (63.4) | 15 (36.6) | |
| History skin tears at 6 months | ||||
| No | 109 | 31 (28.4) | 78 (71.6) | <0.001* |
| Yes | 64 | 40 (62.5) | 24 (37.5) | |
| Density of hair, arms | ||||
| Light | 131 | 45 (34.4) | 86 (65.6) | 0.005** |
| Moderate | 34 | 20 (58.8) | 14 (41.2) | |
| Heavy | 8 | 6 (75.0) | 2 (25.0) | |
| Density of hair, legs | ||||
| Light | 154 | 65 (42.2) | 89 (57.8) | 0.374 |
| Moderate | 19 | 6 (31.6) | 13 (68.4) | |
| History of skin tears at 12 months | ||||
| No | 84 | 19 (22.6) | 65 (77.4) | <0.001* |
| Yes | 89 | 52 (58.4) | 37 (41.6) | |
| Skin characteristics | ||||
| Purpura ≤ 20 mm | ||||
| No | 65 | 11 (16.9) | 54 (83.1) | <0.001* |
| Yes | 108 | 60 (55.6) | 48 (44.4) | |
| Ecchymosis ≥20 mm | ||||
| No | 129 | 46 (35.7) | 83 (64.3) | 0.014* |
| Yes | 44 | 25 (56.8) | 19 (43.2) | |
| Bruising | ||||
| No | 168 | 69 (41.1) | 99 (58.9) | 0.962 |
| Yes | 5 | 2 (40.0) | 3 (60.0) | |
| Haematoma | ||||
| No | 173 | 71 (41.0) | 102 (59.0) | — |
| Yes | 0 | 0.0 (0) | 0.0 (0) | |
| Presence scar tissue | ||||
| No | 56 | 15 (26.8) | 41 (73.2) | 0.008 |
| Yes | 117 | 56 (47.9) | 61 (52.1) | |
| Dermatological skin condition | ||||
| No | 167 | 68 (40.7) | 99 (59.3) | 0.831 |
| Yes | 6 | 3 (50.0) | 3 (50.0) | |
| Lax skin | ||||
| No | 10 | 7 (70.0) | 3 (30.0) | 0.055 |
| Yes | 163 | 64 (39.3) | 99 (60.7) | |
| Fine wrinkles | ||||
| Mild | 91 | 50 (54.9) | 41 (45.1) | <0.001 |
| Moderate/severe | 82 | 21 (25.6) | 61 (74.4) | |
| Coarse wrinkles | ||||
| No | 111 | 42 (37.8) | 69 (62.2) | 0.374 |
| Yes | 62 | 29 (46.8) | 33 (53.2) | |
| Lentigines | ||||
| No | 106 | 44 (41.5) | 62 (58.5) | 0.082 |
| Yes | 67 | 27 (40.3) | 40 (59.7) | |
| Uneven pigmentation | ||||
| No | 62 | 18 (29.0) | 44 (71.0) | 0.016 |
| Yes | 112 | 53 (47.7) | 58 (52.3) | |
| Yellowness | ||||
| No | 169 | 68 (40.2) | 101 (59.8) | 0.162 |
| Yes | 4 | 3 (75.0) | 1 (25.0) | |
| Permanent redness | ||||
| No | 127 | 47 (37.0) | 80 (63.0) | 0.073 |
| Yes | 46 | 24 (52.2) | 22 (47.8) | |
| Elastosis | ||||
| No | 102 | 30 (29.4) | 72 (70.6) | <0.001 |
| Yes | 71 | 41 (57.7) | 30 (42.3) | |
| Pseudoscars | ||||
| No | 161 | 63 (39.1) | 98 (60.9) | 0.061 |
| Yes | 112 | 8 (66.7) | 4 (33.3) | |
| History of actinic keratosis | ||||
| No | 92 | 29 (31.5) | 63 (68.5) | 0.007 |
| Yes | 81 | 42 (51.9) | 39 (48.1) | |
| Malignant skin lesion history | ||||
| No | 122 | 46 (37.7) | 76 (62.3) | 0.168 |
| Yes | 51 | 25 (49.0) | 26 (51.0) | |
| Cutis rhomboidalis nuchae | ||||
| No | 122 | 45 (36.9) | 77 (63.1) | 0.086 |
| Yes | 51 | 26 (51.0) | 25 (49.0) | |
Abbreviations: ADL, activities of daily living; n, number; PAS, psychogeriatric assessment scale; PPG, photoplethysmography; SD, standard deviation; TBPI, toe brachial pressure index.
P‐value from χ 2 test for categorical variables or t‐test for continuous variables
P < 0.05;
P < 0.01.
The baseline characteristics identified the following variables to be significantly associated with the skin tear occurrence at 6 months: age; gender (males); height; principle work environment; respiratory disease; density of hair over the forearms; and a history of falls in the previous 1 month, 3 months, and 6 months and a history of skin tears in the previous 3 months, 6 months, and 12 months. The ability to reposition self, paralysis, and agitation were identified to be significantly associated with not having a skin tear. Skin characteristics found to be significantly associated with skin tears at 6 months included: purpura, ecchymosis, elastosis, and actinic keratosis (AK). The presence of scar tissue, fine wrinkles, and uneven pigmentation was found to be significantly associated with not having a skin tear.
A broad range of variables that were commonly reported in the literature to be associated with skin tears were found not to be significantly associated with the skin tear occurrence at 6 months in participants of this prospective study. These variables included: ethnicity, bruising, cardiac problems, cognitive impairment, contractures, corticosteroid therapy, decreased Braden Scale score, dependency for ADL, haematoma, immobility, oedema, polypharmacy, and application of a moisturiser. Despite a reported link between Caucasian ethnicity and skin tears, the present study could not evaluate this factor because of the relative homogeneity of the ethnicity of the sample in this study. Similarly, in this sample of residents, no association was evident between Fitzpatrick skin types and skin tears, even though participants were classified as having Fitzpatrick skin types that ranged between I and IV.
An evaluation of the difference in mean baseline skin property measurements and transepidermal skin proteins between participants with and without skin tears at 6 months is reported in Table 2.
Table 2.
Independent sample t‐test comparing the means of continuous baseline variables with 6‐month incidence of skin tears
| Variables | Skin tears (n = 71) | No skin tears (n = 102) | P‐value | ||
|---|---|---|---|---|---|
| Mean | SD | Mean | SD | ||
| Skin properties | |||||
| Mean melanin | 29.9 | 3.4 | 29.8 | 4.3 | 0.866 |
| Mean TEWL of forearms | 8.6 | 3.1 | 7.4 | 2.6 | 0.004 |
| Mean TEWL of legs | 5.6 | 1.7 | 5.1 | 1.4 | 0.035 |
| Mean hydration of forearms | 72.7 | 33.0 | 82.0 | 30.2 | 0.058 |
| Mean hydration of legs | 82.6 | 37.3 | 89.3 | 33.6 | 0.217 |
| Mean pH of forearms | 5.8 | 0.5 | 5.6 | 0.5 | 0.004 |
| Mean pH of legs | 5.9 | 0.5 | 5.7 | 0.5 | 0.014 |
| Mean SLEB of forearms | 301.0 | 98.2 | 282.4 | 71.5 | 0.151 |
| Mean SLEB of legs | 161.5 | 69.0 | 133.7 | 70.7 | 0.011 |
| Mean skin thickness of forearms | 829.6 | 193.0 | 821.6 | 172.3 | 0.775 |
| Mean skin thickness of legs | 1075.1 | 261.2 | 1106.9 | 253.2 | 0.424 |
| Mean skin intensity score of forearms | 44.2 | 12.2 | 44.5 | 10.8 | 0.847 |
| Mean skin intensity score of legs | 46.8 | 14.9 | 45.9 | 14.5 | 0.681 |
| Mean VE of forearms | 4.3 | 1.6 | 4.2 | 1.6 | 0.632 |
| Mean VE of legs | 2.6 | 1.3 | 2.4 | 1.1 | 0.177 |
| Mean distensibility of forearms | 13.5 | 2.8 | 13.7 | 2.5 | 0.531 |
| Mean distensibility of legs | 16.4 | 1.9 | 16.6 | 1.8 | 0.515 |
| Mean retraction of forearms | 5.6 | 4.2 | 5.1 | 3.0 | 0.413 |
| Mean retraction of legs | 4.0 | 2.4 | 3.9 | 2.5 | 0.823 |
| Transepidermal skin proteins | |||||
| Mean type IV collagen of forearm | 24.1 | 17.8 | 24.0 | 17.8 | 0.959 |
| Mean type IV collagen of leg | 22.7 | 21.6 | 24.3 | 17.4 | 0.591 |
| Mean MMP‐2 of forearm | 18.9 | 26.8 | 22.1 | 26.7 | 0.428 |
| Mean MMP‐2 of leg | 16.6 | 26.1 | 18.4 | 19.4 | 0.591 |
| Mean TNF‐alpha of forearm | 69.6 | 79.5 | 63.1 | 74.3 | 0.578 |
| Mean TNF‐alpha of leg | 71.9 | 79.7 | 63.3 | 77.5 | 0.481 |
Abbreviations: ADL, activities of daily living; PPG, photoplethysmography; SD, standard deviation; SLEB, subepidermal low echogenicity band; t, t‐test; TBPI, toe brachial pressure index; TEWL, transepidermal water loss; VE, viscoelasticity.
P < 0.05;
P < 0.01.
A statistically significant difference was identified between participants with skin tears and participants without skin tears for the following variables: TEWL dorsal forearms, TEWL legs, pH dorsal forearms, pH legs, and subepidermal low echogenicity band (SLEB) legs.
The findings from the univariable and multivariable analyses are reported in Table 3.
Table 3.
Results of univariable and multivariable analysis of baseline variables associated with skin tears at 6 months
| Variable | Skin tear incidence at 6 months (univariable) | Skin tear incidence at 6 months (multivariable) |
|---|---|---|
| OR (95% CI) | OR (95% CI) | |
| Individual characteristics | ||
| Gender (males versus females) | 3.39 (1.71‐6.71) (P < 0.001) | 3.08 (1.22‐7.77) (P = 0.017) |
| Principle work environment (mix indoor/outdoor versus primarily working outdoors) | 0.20 (0.05‐0.79) (P = 0.021) | |
| History of skin tears in previous 3 months (yes versus no) | 3.35 (1.61‐6.96) (P = 0.001) | |
| History of skin tears in previous 6 months (yes versus no) | 4.19 (2.18‐8.08) (P < 0.001) | |
| History of skin tears in previous 12 months (yes versus no) | 4.81 (2.48‐9.33) (P < 0.001) | 3.82 (1.64‐8.90) (P = 0.002) |
| Falls risk category (medium versus high) | 3.31 (1.75‐6.29) (P < 0.001) | |
| History of falls in previous 1 month (yes versus no) | 2.10 (1.06‐4.15) (P = 0.034) | |
| History of falls in previous 3 months (yes versus no) | 3.69 (1.95‐6.98) (P < 0.001) | 3.37 (1.54‐7.41) (P = 0.002) |
| History of falls in previous 6 months (yes versus no) | 3.45 (1.78‐6.68) (P < 0.001) | |
| Respiratory disease ‐yes versus no | 2.63 (1.03‐6.74) (P = 0.043) | |
| COPD (yes versus no) | 2.54 (1.03‐6.24) (P = 0.043) | |
| Agitation (moderate/severe versus nil/occasionally) | 1.93 (1.03‐3.63) (P = 0.041) | |
| Ability to reposition self (yes versus no) | 1.91 (1.03‐3.54) (P = 0.040) | |
| Skin characteristics | ||
| Purpura/ecchymosis | ||
| Purpura versus none | 5.93 (2.63‐13.37) | |
| Purpura/ecchymosis versus none | 6.46 (2.68‐15.59) (P < 0.005) | |
| Purpura all sites (yes versus no) | 6.14 (2.90‐13.01) (P < 0.001) | 3.64 (1.42‐9.35) (P = 0.007) |
| Ecchymosis all sites (yes versus no) | 2.37 (1.18‐4.77) (P = 0.015) | |
| Scar tissue (yes versus no) | 2.51 (1.25‐5.02) (P = 0 0.009) | |
| Uneven pigmentation (yes versus no) | 2.23 (1.15‐4.33) (P = 0.017) | |
| Fine wrinkles (mild versus moderate/severe) | 0.28 (0.15‐0.54) (P < 0.001) | |
| Actinic keratosis (yes versus no) | 2.34 (1.26‐4.35) (P = 0.007) | |
| Malignant skin lesion (yes versus no) | 2.44 (1.30‐4.59) (P = 0.006) | |
| Cutaneous elastosis (yes versus no) | 3.28 (1.74‐6.19) (P < 0.001) | 3.19 (1.38‐7.38) (P = 0.007) |
| Skin properties and continuous variables (OR based on a unit change in these variables) | ||
| Height | 1.04 (1.01‐1.08) (P = 0.024) | |
| Mean TEWL of bilateral forearms | 1.17 (1.05‐1.31) (P = 0.006) | 1.14 (1.01‐1.28) (P = 0.033) |
| Mean TEWL of bilateral legs | 1.24 (1.01‐1.51) (P = 0.038) | |
| Mean pH of bilateral forearms | 2.57 (1.32‐5.01) (P = 0.006) | 2.56 (1.26‐5.21) (P = 0.010) |
| Mean pH of bilateral legs | 2.34 (1.17‐4.67) (P = 0.016) | |
| Mean SLEB of bilateral forearm | 1.00 (1.00‐1.01) (P = 0.026) | |
| Mean SLEB of bilateral legs | 1.01 (1.00‐1.01) (P = 0.006) | 1.00 (1.00‐1.01) (P = 0.028) |
Abbreviations: CI, confidence interval; OR, odds ratio; SLEB, subepidermal low echogenicity band; TEWL, transepidermal water loss.
Univariable analysis identified 29 variables that were potential predictors of skin tears. Subsequent multivariable analysis identified eight variables that were significantly and independently associated with skin tears at 6 months. These variables comprised three individual characteristics (gender, history of skin tears at 12 months, and history of falls in previous 3 months); two skin characteristics (purpura all sites and cutaneous elastosis); and three skin properties that were significantly associated with skin tears, including increased TEWL of the forearms, increased mean pH of the forearms, and the SLEB across the bilateral lower extremities.
The stepwise method (forward selection) of nominated variables was applied in the multivariable regression analysis. Identified variables that could potentially confound the analysis were accounted for in the statistical model by adjusting for other explanatory variables. These variables included age, gender, history of skin tears, history of falls, and BMI.
The statistical skin tear model that provided the best explanation of the data identified five variables (three individual and two skin characteristics) to be statistically significant predictors of skin tears at 6 months. The results of this analysis are presented in Table 4.
Table 4.
Results of multivariable logistic regression analysis (95% CI) of baseline variables and the risk of skin tears at 6 months in participants (n = 173)
| Variables | Skin tear incidence at 6 months (multivariable) | |
|---|---|---|
| Odds‐ratio (95% CI) | P‐value | |
| Gender | ||
| Females | 1 | |
| Males | 3.08 (1.22‐7.77) | 0.017 |
| History of skin tears in last 12 months | ||
| No | 1 | |
| Yes | 3.82 (1.64‐8.90) | 0.002 |
| History falls in last 3 months | ||
| No | 1 | |
| Yes | 3.37 (1.54‐7.41) | 0.002 |
| Purpura <20 mm | ||
| No | 1 | |
| Yes | 3.64 (1.42‐9.35) | 0.007 |
| Elastosis | ||
| No | 1 | |
| Yes | 3.19 (1.38‐7.38) | 0.007 |
The individual and skin characteristics that were identified to be significantly associated with skin tears at 6 months included: male gender, history of skin tears in the previous 12 months, history of falls within the preceding 3 months, cutaneous manifestations of elastosis, and purpura. The analysis did not identify any morphological (colour, thickness, elasticity) or physiological (TEWL, hydration, pH) skin property that significantly predicted the risk of skin tears at 6 months.
An ROC curve was generated to assess the performance of the skin tear model (Figure 1).
Figure 1.
Receiver operating characteristic (ROC) curve showing the sensitivity and specificity of the model for various cut‐off values to determine individuals at risk of skin tears [Colour figure can be viewed at wileyonlinelibrary.com]
The model yielded an AUC of 0.854, which indicates that the skin tear model provides “very good discrimination” to correctly classify participants with and without skin tears. The skin tear model correctly predicted 81.7% (sensitivity) of participants with skin tears and 81.4% (specificity) of participants without skin tears.
4. DISCUSSION
The objective of this study was to construct a skin tear predictive model to identify older Australians at risk of these wounds. Multivariable logistic analysis identified male gender, two adverse events (previous history of skin tears and a history of falls), and two clinical manifestations (clinical purpura and elastosis) to significantly predict the risk of skin tears in participants.
In this study population, males were three times more likely to develop a skin tear when compared with females. This result is similar to findings from a point‐prevalence survey of skin tears conducted in a long‐term care facility in Canada on 113 residents.8 Two other studies, undertaken using long‐term elderly Japanese residents, did not identify any significant difference between gender and the occurrence of skin tears.9, 13 In male participants with skin tears, the majority of injuries occurred on the dorsal forearms, particularly the right arm. Male participants in this study typically reported experiencing higher amounts of exposure to UV radiation in the course of their work and recreational life than females. Moreover, motor vehicles are driven on the left side of the road in Australia, where the right forearm is subjected to chronic sun exposure and the influences of photoageing from the penetrating effects of UV radiation.32
The model identified two adverse events, including a previous history of skin tears and a previous history of falls, that predicted the risk of skin tears. Participants with a history of skin tears were nearly four times more likely to develop a skin tear at 6 months. Three previous studies reported similar outcomes, but variation in their research design and the limited quantitative analysis did not permit further evaluation.12, 33, 34 LeBlanc et al.8 suggested that a possible relationship existed between having a history of skin tears and the occurrence of skin tears. Further comparison with this study was limited by its point‐prevalence research design, which recorded only a limited amount of information. A previous history of skin tears suggests that underlying changes to the biomechanical properties of skin leave it vulnerable to repeated trauma‐related injuries.
Similarly, participants with a history of falls in the previous 3 months were more than three times more likely to develop a skin tear at 6 months. The identification of falls as an independent predictor of skin tears is perhaps not surprising as numerous studies of falls in older individuals show that skin tears are a commonly reported injury that occur from this adverse event.35, 36, 37, 38, 39 Not unlike a history of skin tears, the association between falls and skin tears suggests that, in ageing skin, underlying structural changes predispose some participants to increased risk of trauma‐related skin injuries.
The statistical model identified two clinical characteristics—cutaneous manifestations of purpura and elastosis—to be approximately three and three and a half times, respectively, more likely to be associated with an increased risk of skin tears. Purpura, in this study, referred to the clinical manifestation of non‐inflammatory, non‐palpable, ecchymotic lesions on the skin that ranged in diameter from 2 to 20 mm. The presence of purpura in this study was the result of age‐related skin changes and not from any underlying medical conditions. Purpura has previously been reported to be associated with skin tears.3 Decreased amounts of collagen that support small blood vessels are reported to contribute to the risk of purpura.40, 41
Cutaneous manifestations of elastosis in this study referred to coarse, thickened, scaly, dry, and rigid textured skin characteristics that occurred across exposed skin surfaces.42, 43 Identification of clinical elastosis was made by comparing the texture of exposed skin surface with non‐exposed or minimally exposed skin sites, such as an adjacent skin site. No previous skin tear study has reported textural skin changes from cutaneous manifestations of elastosis to be an independent risk factor for skin tears. Earlier work by Koyano et al., 13 demonstrated that solar elastotic changes in the dermis, identified using 20‐MHz ultrasonography to measure the SLEB, was a potential risk factor for skin tears. Thickening of the dermis with concomitant loss of elasticity is reported to increase the risk of skin tears.44
The upper extremities, together with the face and neck, are anatomical sites with the greatest potential for exposure to UV radiation.45 Vertical skin surfaces of an erect person have been reported to receive about 50% of ambient UV radiation, while horizontal surfaces receive as much as 75%.46 It is possible that the dorsal surface of the upper extremities is also exposed to similar levels of UV exposure. While exposure of skin to UV radiation promotes Vitamin D synthesis, chronic exposure causes photoage‐related degenerative changes that impact cellular, fibrous, and vascular skin structures.47, 48, 49, 50 The cumulative effects of UV radiation on exposed skin surfaces have been shown to clinically manifest as elastotic skin changes, purpura, uneven skin pigmentation and melanoma and non‐melanoma skin lesions.51, 52, 53, 54, 55, 56, 57, 58
Despite a broad range of individual attributes, skin characteristics, and skin properties being assessed, this study may not have identified all factors that could contribute to the risk of skin tears. Therefore, the results of this study may only be relevant to Australia, with further studies across different population groups and geographical locations needed to confirm these findings. To better understand the inclusion of the two clinical characteristics (cutaneous manifestations of elastosis and purpura) into the model and the associated risk of skin tears, further analysis was conducted, which will be discussed in greater detail in a subsequent publication using identical data by the same authors.
5. CONCLUSION
This skin tear risk predictive model showed very good discriminative ability to predict the risk of skin tears at 6 months in older Australians. The predictive model identified five variables (male gender, history of skin tears, history of falls, clinical elastosis, and purpura) that were factors for at‐risk persons. These predictors can be readily assessed in any clinical setting and by all health care providers, regardless of their level of training or experience. The model has the potential to guide clinical decision‐making and more efficient targeting of preventive strategies to achieve better health outcomes in terms of reducing the incidence of skin tears and decreasing the costs and time associated with the provision of wound care.
ACKNOWLEDGEMENTS
Robyn Rayner was a recipient of a 2013 Australian Postgraduate Award and Curtin University Postgraduate Scholarship. The authors acknowledge the support of the Wound Management Innovation Cooperative Research Centre and the Australian Government's Cooperative Research Centres Program. The authors are indebted to the residents and staff of the Bethanie Group Inc for their support in undertaking this study. The authors also offer their sincere gratitude to Professor Hiromi Sanada and her research team from the Department of Gerontological Nursing/Wound Care Management at the University of Tokyo for their collaboration in undertaking this research. They are particularly grateful to Dr Takeo Minematsu for assisting them with skin blotting investigations and analysis.
Rayner R, Carville K, Leslie G, Dhaliwal SS. A risk model for the prediction of skin tears in aged care residents: A prospective cohort study. Int Wound J. 2019;16:52–63. 10.1111/iwj.12985
REFERENCES
- 1. Everett S, Powell T. Skin tears—the underestimated wound. Prim Intent. 1994;2:28‐31. [Google Scholar]
- 2. Malone ML, Rozario N, Gavinski M, Goodwin J. The epidemiology of skin tears in the institutionalized elderly. J Am Geriatr Soc. 1991;39:591‐595. [DOI] [PubMed] [Google Scholar]
- 3. White MW, Karam S, Cowell B. Skin tears in frail elders: a practical approach to prevention. Geriatr Nurs. 1994;15:95‐99. [DOI] [PubMed] [Google Scholar]
- 4. Carville K, Lewin G, Newall N, et al. STAR: a consensus for skin tear classification. Prim Intent. 2007;15:18‐28. [Google Scholar]
- 5. Ratliff C, Fletcher KR. Skin tears: a review of the evidence to support prevention and treatment. Ostomy Wound Manage. 2007;53:32‐42. [PubMed] [Google Scholar]
- 6. Strazzieri‐Pulido KC, Peres GRP, Campanili TCGF, de Gouveia Santos VLC. Incidence of skin tears and risk factors: a systematic literature review. J Wound Ostomy Cont Nurs. 2017;44:29‐33. [DOI] [PubMed] [Google Scholar]
- 7. Payne R, Martin M. Defining and classifying skin tears: need for a common language. Ostomy Wound Manage. 1993;39:16‐26. [PubMed] [Google Scholar]
- 8. LeBlanc KA, Christensen D, Cook J, Culhane B, Gutierrez O. Prevalence of skin tears in a long‐term care facility. J Wound Ostomy Continence Nurs. 2013;40:1‐5. [DOI] [PubMed] [Google Scholar]
- 9. Sanada H, Nakagami G, Koyano Y, Iizaka S, Sugama J. Incidence of skin tears in the extremities among elderly patients at a long‐term medical facility in Japan: a prospective cohort study. Geriatr Gerontol Int. 2015;15:1058‐1063. [DOI] [PubMed] [Google Scholar]
- 10. Koyano Y, Nakagami G, Iizaka S, Sugama J, Sanada H. Skin property can predict the development of skin tears among elderly patients: a prospective cohort study. Int Wound J. 2017;14:691‐697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Carville K, Leslie G, Osseiran‐Moisson R, Newall N, Lewin G. The effectiveness of a twice‐daily skin‐moisturising regimen for reducing the incidence of skin tears. Int Wound J. 2014;11:446‐453. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Payne R, Martin M. The epidemiology and management of skin tears in older adults. Ostomy Wound Manage. 1990;26:26‐37. [PubMed] [Google Scholar]
- 13. Koyano Y, Nakagami G, Iizaka S, et al. Exploring the prevalence of skin tears and skin properties related to skin tears in elderly patients at a long‐term medical facility in Japan. Int Wound J. 2014;13:189‐197. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Rayner R, Carville K, Leslie G, Roberts P. A review of patient and skin characteristics associated with skin tears. J Wound Care. 2015;24:406‐414. [DOI] [PubMed] [Google Scholar]
- 15. Lewin GF, Newall N, Alan JJ, Carville KJ, Santamaria NM, Roberts PA. Identification of risk factors associated with the development of skin tears in hospitalised older persons: a case‐control study. Int Wound J. 2016;13:1246‐1251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Newall N, Lewin GF, Bulsara MK, Carville KJ, Leslie GD, Roberts PA. The development and testing of a skin tear risk assessment tool. Int Wound J. 2016;14:97‐103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Rayner R, Carville K, Leslie G, Dhaliwal SS. Skin tear risk associated with aged care residents: a cross‐sectional study. Wound Pract Res. in press. [Google Scholar]
- 18. Rayner R, Carville K, Leslie G, Dhaliwal SS. Measurement of morphological and physiological skin properties in aged care residents: a test–retest reliability pilot study. Int Wound J. 2017;14:420‐429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. World Medical Association . World medical association declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013;310:2191‐2194. [DOI] [PubMed] [Google Scholar]
- 20. Australian Government National Health and Medical Research Council . Australian code for the responsible conduct of research. Australian Government; 2007:1–12.3.
- 21. Minematsu T, Horii M, Oe M, et al. Skin blotting: a noninvasive technique for evaluating physiological skin status. Adv Skin Wound Care. 2014;27:272‐279. [DOI] [PubMed] [Google Scholar]
- 22. Hertzog MA. Considerations in determining sample size for pilot studies. Res Nurs Health. 2008;31:180‐191. [DOI] [PubMed] [Google Scholar]
- 23. Dawson B, Trapp RG. Basic & Clinical Biostatistics. 3rd ed. Boston, MA: Lange Medical Books/McGraw‐Hill Medical Publishing Division; 2001. [Google Scholar]
- 24. Peat JK, Mellis C, Williams KA, Xuan W. Calculating the Sample Size. Health Science Research: A Handbook of Quantitative Methods. London, England: Sage; 2002:128‐147. [Google Scholar]
- 25. Flament F, Bazin R, Laquieze S, Rubert V, Simonpietri E, Piot B. Effect of the sun on visible clinical signs of aging in Caucasian skin. Clin Cosmet Investig Dermatol. 2013;6:221‐232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. MJL P. Chi‐square. In: Kreutzer J, DeLuca J, Caplan B, eds. Encyclopedia of Clinical Neuropsychology. Cham, Switzerland: Springer International Publishing; 2017:1. [Google Scholar]
- 27. Tsai AC. Achieving consensus on terminology describing multivariable analyses. Am J Public Health. 2013;103:e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Wakkee M, Hollestein LM, Nijsten T. Multivariable analysis. J Invest Dermatol. 2014;134:1‐5. [DOI] [PubMed] [Google Scholar]
- 29. Szumilas M. Explaining odds ratios. J Can Acad Child Adolesc Psychiatry. 2010;19:227‐229. [PMC free article] [PubMed] [Google Scholar]
- 30. Florkowski CM. Sensitivity, specificity, receiver‐operating characteristic (ROC) curves and likelihood ratios: communicating the performance of diagnostic tests. Clin Biochem Rev. 2008;29:S83‐S87. [PMC free article] [PubMed] [Google Scholar]
- 31. Meyers LS, Gamst GC, Guarino A. Performing Data Analysis Using IBM SPSS. Hoboken, NJ: John Wiley & Sons; 2013. [Google Scholar]
- 32. Lewis KG, Bercovitch L, Dill SW, Robinson‐Bostom L. Acquired disorders of elastic tissue: part I. Increased elastic tissue and solar elastotic syndromes. J Am Acad Dermatol. 2004;51:1‐21. [DOI] [PubMed] [Google Scholar]
- 33. McGough‐Csarny J, Kopac C. Skin tears in institutionalized elderly: an epidemiological study. Ostomy Wound Manage. 1998;44:14‐25. [PubMed] [Google Scholar]
- 34. Skiveren J, Wahlers B, Bermark S. Prevalence of skin tears in the extremities among elderly residents at a nursing home in Denmark. J Wound Care. 2017;26:S32‐S36. [DOI] [PubMed] [Google Scholar]
- 35. Butler M, Kerse N, Todd M. Circumstances and consequences of falls in residential care: the New Zealand story. NZ Med J. 2004;117:1076. [PubMed] [Google Scholar]
- 36. Resnick B, Junlapeeya P. Falls in a community of older adults: findings and implications for practice. Appl Nurs Res. 2004;17:81‐91. [DOI] [PubMed] [Google Scholar]
- 37. Chang C‐M, Lin H‐F, Chiang H‐H. A study on the relationship between age and inpatient falls in Taiwan. Int J Nurs Pract. 2015;21:605‐611. [DOI] [PubMed] [Google Scholar]
- 38. Krauss MJ, Evanoff B, Hitcho E, et al. A case‐control study of patient, medication, and care‐related risk factors for inpatient falls. J Gen Intern Med. 2005;20:116‐122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Bank D, Nix D. Preventing skin tears in a nursing and rehabilitation center: an interdisciplinary effort. Ostomy Wound Manage. 2006;52:38‐46. [PubMed] [Google Scholar]
- 40. Leo MS, Sivamani RK. Cosmeceutical Treatments for Purpura. Cosmeceuticals and Active Cosmetics. 3rd ed. Boca Raton, FL: CRC Press; 2015:365‐368. [Google Scholar]
- 41. Gloster HM Jr, Gebauer LE, Mistur RL. Photodermatoses. Absolute Dermatology Review. Cham, Switzerland: Springer; 2016:423‐427. [Google Scholar]
- 42. Patterson JW. Disorders of elastic tissue. Weedon's Skin Pathology. 4th ed. New York, USA: Elsevier Health Sciences; 2016:381‐403. [Google Scholar]
- 43. Raimer SS, Sanchez RL, Hubler WR Jr, Dodson RF. Solar elastotic bands of the forearm: an unusual clinical presentation of actinic elastosis. J Am Acad Dermatol. 1986;15:650‐656. [DOI] [PubMed] [Google Scholar]
- 44. Ling LC. Aging in Asian skin. In: Farage MA, Miller KW, Maibach HI, eds. Textbook of Aging Skin. Berlin Heidelberg, Germany: Springer‐Verlag; 2010:1019‐1024. [Google Scholar]
- 45. Pastila R. Effects of ultraviolet radiation on skin cell proteomeIn: Leszczynski D, ed. Radiation Proteomics: The Effects of Ionizing and Non‐Ionizing Radiation on Cells and Tissues. Dordrecht, the Netherlands: Springer Netherlands; 2013:121‐127. [Google Scholar]
- 46. Diffey BL. Solar ultraviolet radiation effects on biological systems. Phys Med Biol. 1991;36:299‐328. [DOI] [PubMed] [Google Scholar]
- 47. Guimarães C, Bagatin E, Guadanhim L, et al. Development and validation of a clinical scale for the evaluation of forearm skin photoaging. J Cutan Med Surg. 2015;19:380‐387. [DOI] [PubMed] [Google Scholar]
- 48. Calderone DC, Fenske NA. The clinical spectrum of actinic elastosis. J Am Acad Dermatol. 1995;32:1016‐1024. [DOI] [PubMed] [Google Scholar]
- 49. Caetano L, Soares JLM, Bagatin E, Miot HA. Reliable assessment of forearm photoageing by high‐frequency ultrasound: a cross‐sectional study. Int J Cosmet Sci. 2015;38:170‐177. [DOI] [PubMed] [Google Scholar]
- 50. World Health Organisation . Global Solar UV Index. A Practical Guide. WHO/SDE/OEH/022. Geneva, Switzerland: World Health Organisation; 2002:1‐28. [Google Scholar]
- 51. Friedman BJ, Lim HW, Wang SQ. Photoprotection and photoaging. In: Wang QS, Lim WH, eds. Principles and Practice of Photoprotection. Berlin Heidelberg, Germany: Springer International Publishing; 2016:61‐74. [Google Scholar]
- 52. Yin R, Chen Q, Hamblin MR. Skin Aging and Photoaging. San Rafael, CA: Morgan & Claypool Publishers; 2015:1‐4. [Google Scholar]
- 53. Rittié L, Fisher GJ. Natural and sun‐induced aging of human skin. Cold Spring Harb Perspect Med. 2015;5:a015370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Green AC, Hughes MCB, McBride P, Fourtanier A. Factors associated with premature skin aging (photoaging) before the age of 55: a population‐based study. Dermatology. 2011;222:74‐80. [DOI] [PubMed] [Google Scholar]
- 55. Biniek K, Levi K, Dauskardt RH. Solar UV radiation reduces the barrier function of human skin. Proc Natl Acad Sci. 2012;109:17111‐17116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56. Oba A, Edwards C. Relationships between changes in mechanical properties of the skin, wrinkling, and destruction of dermal collagen fiber bundles caused by photoaging. Skin Res Technol. 2006;12:283‐288. [DOI] [PubMed] [Google Scholar]
- 57. Varani J, Dame MK, Rittie L, et al. Decreased collagen production in chronologically aged skin. Roles of age‐dependent alteration in fibroblast function and defective mechanical stimulation. Am J Pathol. 2006;168:1861‐1868. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58. Young C. Solar ultraviolet radiation and skin cancer. Occup Med. 2009;59:82‐88. [DOI] [PubMed] [Google Scholar]