Arica is a city located in a subtropical area in the northernmost region of Chile. Arica has a microclimate with fairly stable meteorological conditions throughout the year, including a lack of rainfall (<5 mm a decade), predictable winds and a high percentage of sunny days per year (15,16). In addition, Arica is a desert area with vast sand and sea surfaces as well as scarce vegetation; therefore, Arica has high ground reflectivity (14,15). The city is located at sea level and due to the good climate conditions, the population perform numerous outdoor recreational activities. (16,17).

A UVB sun device, which has been in used at the UV Sunlight Radiation Laboratory of Tarapacá University (Arica, Chile) since September 2006, was used in the present study. It is located 25 m above sea level, at a latitude of 18° 49′S and longitude of 70° 19′W on the uppermost part of the building with no elements blocking incoming sunlight radiation. Erythemal UV irradiance (UV_Ery) was measured using a calibrated high-range YES-UVB-1 biometer (serial no. 090710, Yankee Environmental Systems, Inc., Turners Falls, MA, USA) through an agreement between Tarapacá University and the Chilean Meteorological Department (CMD, Santiago, Chile). Measurements obtained from the device were incorporated in real-time using a UVB device network installed by the CMD. Thus, the device was connected to the UVB sun-measuring network of Chile, covering an area between Arica and the Antarctic territory in the South of Chile. All devices used in the present study complied with the World Meteorological Organization calibration specifications (18).

The solar UVI (W/m²) was calculated from the collected UV_Ery biometer data by multiplying UV_Ery by 40 (11). Subsequently, the T_Ery [1 minimal erythemal dose (MED)] was calculated using the following equation: T_Ery = (4,000/60) × [(MED × sun protection factor)/UVI)], where MEDF depends on the Fitzpatrick classification of skin types (19,20). In the present study, the cohort included individuals with MED skin type IV (0.40–0.60 kJ/m²/day). The cohort used in this study has been further described previously (15).

UV_Ery is expressed as the standard erythemal dose (SED) required to induce erythema, where 1 SED=100 Jm⁻² UV_Ery. From the UV_Ery biometer measurement, vitamin D UV irradiation (UV_VitD) was estimated using a tropospheric UV model (21). Ozone layer data were obtained from NASA/TOMS measurements by using the User's Guide from National Aeronautics and Space Administration (NASA) (22). The subsequent calculation for TVitD was obtained from Mckenzie et al (3). The most common skin types in chilean teenagers of low social economical strata that have been analyzed in previous studies (14–16) were used in the present study. In this study, the cohort included 98 patients treated at the Pathology Department of Dr Juan Noé Crevani Hospital (Arica, Chile) between January 1989 and December 1993 (14). Furthermore, when ozone layer and solar zenith angle (SZA) data were known, the T_VitD was calculated using the following formula: T_VitD= 20/(UVI × R), where R is a quotient of UV_VitD/UV_Ery. The ratio function of UV_VitD and UV_Ery was adjusted using the following equation: UV_VitD/UV_Ery = (2.00524–1.48303) × exp (−UVI/2.73854). In addition, the skin cancer rates (SCR) per 100,000 individuals between 2001 and 2011 (14,23) were adjusted using the following linear fit equation: SCR = (−918.24 + 1.46) × number of years.

The study was approved by the ethics committee of Servicio de Dermatología, Dr Juan Noé Crevani Hospital (Arica, Chile). The data analyzed in the study were obtained from public databases of Servicio de Dermatología, Dr Juan Noé Crevani Hospital.

The present study investigated the association between the risk of skin cancer and mild exposure to UV light in Arica, Chile. The T_Ery and T_VitD were calculated to explain the possible effects of insufficient vitamin D in individuals living at a latitude of 18° 18′S, a longitude of 70°, 26′W and a height of 23 m. Fig. 1A displays the highest daily levels of solar UVI versus Julian day (in one year Julian days take the values 1–365), the second scale indicate the years. Solar UVI values were obtained in Arica between September 20, 2006 and November 24, 2011, with the solar UVI fluctuating between 2 and 15.5 W/m². Fig. 1B indicates the quotients of UV_VitD/UV_Ery as a function of UVI and demonstrates that the quotient varied between 0.5 and 2.0 for UVI values between 0 and 15 W/m².

Fig. 2A reveals the results obtained from calculating the T_Ery in the four different seasons of the year between 2007 and 2013. It was observed that the T_Ery varied between 12 min in the summer and 34 min in the winter for skin type IV. Fig. 2B displays the mean value T_VitD per person during the four seasons of the year when only hands and face were directly exposed to UV sunlight (10% of the surface area of the body). These calculations were obtained between 2007 and 2013, and varied from 8 min in the summer to 24 min in the winter.

Fig. 3A indicates the average monthly T_VitD required for individuals that only exposed their hands and face to direct sunlight. These calculations were made at midday each month between 2006 and 2009. It was observed that the exposure times varied between 8 min in summer (January-February) and 24 min in winter (June-July). In addition, the calculated T_VitD data was similar each year. Fig. 3B shows the SCR per 100,000 individuals between 2001 and 2006 (23), and 2007 and 2011 (14). A steady increase was observed, increasing from 6.9 in 2001 to 23.8 in 2011, with the highest rate reaching 30.8 in 2009.

The quotient of UV_VitD/UV_Ery as a function of the SZA is displayed in Fig. 4A. This calculation was performed using UVI data obtained in 2007, which behaved similarly to previous years, using an ozone layer thickness of 250 Dobson units (DU). The highest value for this quotient (2.06–2.07) was obtained at a SZA of <10°, whereas the lowest value for this quotient (1.83) was obtained at a SZA of >50°. The quotient of UV_VitD/UV_Ery as a function of UVI data from 2007, with an ozone layer thickness of 250 DU is indicated in Fig. 4B. The highest value for this quotient (2.06–2.07) had a UVI value of >10 W/m² and the lowest value for this quotient (1.83) had a UVI of 5 W/m². Fig. 4C demonstrates solar UVI as a function of SZA between January and July 2007. The highest UVI value was 12.7 W/m² for a SZA that fluctuated between 5 and 20°, with a maximum and minimum value of 5 and 40°, respectively. Table I presents the skin cancer rate and time to synthesize 1000 UI of vitamin D (TVID) during the four seasons of the year between 2007 and 2011 in Arica, Chile. The mean SCR between 2007 and 2011 was calculated as 26.9 per 100,000 individuals.