The number of common, benign pigmented moles a person develops is the strongest marker of melanoma risk [5], with the presence of multiple moles increasing the risk of melanoma 20-fold [6]. This association has been verified in numerous case-control studies [5]. Only one percent of babies are born with a pigmented mole, yet many people develop tens to hundreds of them by adulthood [5], with the highest densities occurring on sun-exposed body-sites [7].

Excessive mole development is related to childhood sun exposure [8]. Caucasian immigrants who arrive in Australia after 10 years of age develop fewer pigmented moles than the “native” Caucasian population and are less likely to develop melanoma [9] [10]. Likewise, children from tropical Australia develop pigmented moles earlier and in higher numbers than children from temperate regions in Australia or abroad, and are at greater risk of melanoma as a consequence [11] [12]. The association between sun exposure and mole development includes acute and chronic patterns of exposure, with a single sunburn before age 7 years almost doubling the risk of excessive mole development, and spending four or more hours outdoors each day more than tripling the risk [8].

Sun-protective clothing, such as broad-brim hats and long-sleeved shirts provide a physical barrier that helps to protect the skin from UVR [13]. Swimsuits that cover the trunk and incorporate longer sleeves and pants are ideal for water-based activities since, unlike sunscreen, they do not require reapplication [14] [15]. Ecological evidence suggests that Australian children born since Nylon Elastane rash-vests and all-in-one sun-suits gained popularity, have fewer pigmented moles on their back than children born earlier [16]. More recently, the only randomized controlled trial to evaluate the efficacy of sun-protective clothing confirmed that wearing garments made from darker-colored, tightly woven fabrics covering a substantial proportion of the body slows the rate of mole development in Caucasian Australian children [17]. Thus, well designed sun-protective clothing can reduce melanoma risk in susceptible populations, particularly if worn regularly from early childhood [17].

In July 1996, Australia was the first nation to introduce a standard for the evaluation and classification of clothing claiming a sun-protective advantage [18]. Known as AS/NZS 4399:1996, this joint Australian and New Zealand Standard introduced a reproducible measurement and classification protocol based on the relative ranking of the transmission of UVR through the fabric from which the garment is made [19]. This standard outlines all of the technical specifications related to measuring UVR transmittance as well as describing the Ultraviolet Protection Factor (UPF) labeling specifications that manufacturers, wholesalers and retailers wishing to classify garments according to the UPF classification scheme are expected to adhere to [19].

The UPF-labelling of sun-protective garments was devised with the intention of enabling consumers to make informed clothing purchases. The standard and its associated UPF rating system is still widely used today, almost 20 years since it was introduced [18]. Fabrics that achieve UPF ratings of 40, 45, 50 or 50+ block at least 97.5% of erythemally effective (EE) UVR (i.e. those wavelengths in the ultraviolet spectrum responsible for causing the sunburn response in human skin). In simple terms, a UPF 50 fabric allows only 1/50^th of the UVR to reach the skin [18]. Fabrics with UPF ratings of 40 and above are considered to provide “excellent” sun-protec- tion, and can be labelled as such, while fabrics in the “very good” category have UPF ratings of 25 - 35 and block 96% - 97.5% of EE-UVR [20]. Similarly, fabric bolts and apparel rated as UPF 15 - 20 block 93.3% - 95.9% of EE-UVR and can display the words “good sun-protection” on their swing tags [18]-[20].

AS/NZ 4399:1996 and its associated UPF rating system have been adopted almost universally by the textile industry. However, it only evaluates the transmission of EE-UVR through fabric, without taking into consideration the proportion of the body surface coverage afforded by the design of the garment [13]. While ever AS/NZS 4399:1996 does not specify a minimum body surface proportion that must be covered by a garment in order to claim a sun-protective advantage it will be vulnerable to exploitation. For instance, a sun-protective advantage could be claimed for brief or scant items of apparel, such as bikini swimwear, because they are made of fabrics such as Nylon Elastane (e.g. LYCRA^®) that often achieves a 50+ UPF rating [14]. Under a standard where UPF is the only determinant used to assess the sun-protective benefit of clothing [19], brief swimwear typically achieves the technical specifications of the “excellent” sun-protection category [13]. Thus, currently AS/NZS 4399:1996 provides an avenue whereby manufacturers of brief swimwear and apparel could feasibly make misleading claims of a sun-protective advantage. This loop-hole needs to be closed [13].

In light of recent evidence that garments that cover a greater proportion of the body’s surface area (BSA) help to protect children from developing numerous pigmented moles [17], plans to develop a rating system that evaluates both the BSA and UPF of sun-protective garments seems warranted [13] [17]. Standards Australia and Standards New Zealand Technical Committee TX-021 for the Evaluation and Classification of Sun Protective Clothing recently convened to revise AS/NZ 4399:1996 to address the potential vulnerabilities that have come to light during the 19 years this ground-breaking Standard has been in use [21]. Public comments on proposed revisions to the Standard are invited from industry to ensure it will represents the views of all stakeholders [21].

Here, we discuss our research into and progress towards developing an evaluation protocol devised with the specific intention of quantifying the BSA coverage of sun-protective garments.

The felt covered surface of a flexible mannequin was securely fastened to a typical optical bench in our Physics laboratory using a clamp around the neck and the left ankle, such that it was lying parallel to the bench. The mannequin was marked with horizontal bands drawn with a red indelible marker at 1 cm intervals with the aid of a Bosch laser level on a custom tripod. After the mannequin was marked up into cross-sections in this way from the top of the neck, through the torso and the full length of the right arm and the right lower limb, it was returned to the upright position, supported by its stand. This process was then repeated for mannequins of all sizes.

The first mannequin was placed on top of a table, against a wall covered with a neutral (black) non-reflective background. This was achieved using a backdrop of black felt. Photography lamps were positioned to prevent shadows from obstructing the view of the red lines marked on the mannequin as it was imperative that the lines be clearly visible in the digital images. An 8.0 Megapixel Canon digital SLR camera, fitted with an 18 to 55 mm zoom lens at f4.5 and supported by a tripod, was positioned a fixed distance from the mannequin and orientated with respect to a reference point, marked on the trunk of the mannequin and collinear with the camera’s optical axis. The mannequin image was framed from head to toe using the zoom function of the fitted camera lens to ensure the highest possible image resolution of the marks on the mannequin’s surface. A photograph of the unclothed mannequin was taken using the camera’s inbuilt fill-in flash. The image of the unclothed mannequin is shown in Figure 1 (at left). The same mannequin was then clothed in an upper body garment of the matching size and re-photographed (see Figure 2, at left). This process was repeated for mannequins of all sizes.

In the interest of being able to develop a rating system to inform consumers about the relative BSA covered by separate garments, it is necessary to be able to calculate GPF for both upper and lower body garments. For consistency with the newest iteration of the Standard, the definitions of the upper and the lower body regions used in this protocol match those used in the draft revision DR AS/NZ 4399:2015 [21]. The upper body is defined as including the anterior, posterior and lateral aspects of the torso extending upwards from the hip line to the base of the neck and across the shoulders and down the arms to the wrist. Upper body garments include items such as rash-vests, polo shirts, button-up shirts, and t-shirts. As it would be misleading to permit brief garments to claim a sun-protective advantage [13], in the proposed rating system, a claim of sun-protective advantage will be indicated by superior GPF and UPF values, and will only be granted to and displayed by upper body garments that fully cover the shoulders and from the neck down to the hip line, so that the posterior torso is fully covered. This definition allows for the anterior surface of the garment to open partially or fully (via buttons or a zipper etc.).

Consistent with the draft Standard, the lower body is defined as extending from the hip line, down the legs to the ankles, such that the area considered incorporates the buttocks, the pubic region, and the anterior, posterior and lateral aspects of the thighs and the lower legs [21]. Lower body garments include trousers, shorts, skorts, culottes, and skirts. Lower body GPF will not be displayed on garments such as swimwear that does not cover to at least the midway point of the crotch and the knees to prevent manufacturers and retailers from inappropriately claiming a sun-protective advantage for brief lower body garments [21].

All lines visible above the hip line and on the arms and neck of the mannequin clothed with an upper body garment are counted to determine GPF as described in Equation (1) above. While on a mannequin clothed with a lower body garment, all lines visible from the hip line down to the ankle are counted to determine GPF with the same equation. Further configurations will cater for all-in-one garments like Nylon Elastane sun-suits.

To expedite the calculation of GPF for either upper body or lower body garments, an algorithm was developed using MATLAB software Release 2013b [24] to count the number of lines marked on the “unclothed” and “clothed” mannequin models. Digital images of the “clothed” and “unclothed” mannequins (taken in accordance with the photography protocol described in 3.2) were imported into MATLAB. The mannequin image was then framed relative to the background to outline the body surface area. This was achieved by removing all image pixels with a red, green and blue saturation of less than 0.75 on a scale of 0 to 1. Marked line segments on the mannequin model were isolated by comparison of the difference in red and green, and red and blue levels in each image pixel defined to be on the mannequin surface. In this way, the marked surface lines, (drawn on the mannequin in red) were classified by the algorithm for pixels which had more red than green, and also more red than blue.

The difference used to classify a marked line on the mannequin model is variable and can be accommodated for by the algorithm by enhancing image brightness and contrast. Line marks automatically classified by the algorithm were given a false color and counted by the algorithm for either the upper body or lower body sections (the upper and lower body sections are divided anatomically at the hip line as described in 4.1). An example of the algorithm is shown for an upper body garment in Figure 2.