Chapter 1: Damage from Acute vs Chronic Solar Exposure

The acute and chronic effects of solar ultraviolet radiation (UVR) exposure are well established. The acute effects can be readily studied under controlled laboratory conditions whereas the chronic effects have been determined by epidemiology in the case of skin cancer and by clinical observation in the case of photoageing. Chronic exposure is, by definition, a series of acute (i.e. single) exposures; however, the relationship between photodamage by acute and chronic exposure is very poorly understood, mainly because this has not been extensively studied. The patterns of exposure seem to be important in long-term clinical outcome. Epidemiological research suggests that regular exposure is important in squamous cell carcinoma (SCC) [1], which is often associated with signs of photoaging such as elastosis. In contrast, intermittent sunburning exposure seems to be important in malignant melanoma (MM) [1,2] as does childhood exposure. There is also evidence that intermittent exposure is important in basal cell carcinoma (BCC) [1,3].

Most UVR exposure in healthy people comes from the sun, but in recent years exposure from tanning devices has become increasingly widespread, especially amongst young women, and this has raised concern about long-term adverse effects such as MM and, to a lesser extent, SCC [4].

Controlled chronic UVR exposure to clinical outcome, such as skin cancer or photoageing, is neither ethical nor practicable in humans, but it is possible to carry out studies with repeated UVR exposure over relatively short periods. This mimics “real life” more realistically than acute exposure experiments and also allows the study of adaptive responses that may influence the response to subsequent exposure. Furthermore, repeat exposure studies act as a bridge between the acute and the chronic effects of UVR and may provide a better understanding of how individual exposures result in long-term clinical outcome.

The discussion in this chapter will focus on the acute effects of physiologically and environmentally relevant UVR exposure on human skin and how these effects may be modified by repeated exposure. In addition, the possible effects of repeated exposure on long-term clinical outcome will be considered. The relationship between sunscreen photoprotection of acute and chronic photodamage will also be discussed.

Erythema (inflammation) is the most obvious clinical sign of UVR exposure and is apparent from about 6 hours after exposure and is maximal at about 24 hours [5]. Its action spectrum is maximal at about 300 nm, which is about three orders of magnitude more effective than UVA [6,7]. This peak has been made into a plateau in the widely used mathematically derived curve known as the Commission Internationale de l′Éclairage (CIE) erythema reference action spectrum [8]. The minimal erythema dose (MED) is used as a means of defining personal sensitivity to UVR, and is defined as the UVR dose (J/m²), of a given spectrum, that causes a (just) perceptible skin reddening. In other words, the MED is the visual threshold of UVR dose-response of which erythema becomes more intense with higher doses. This can be demonstrated by visual grading or by the measure of redness by reflectance spectroscopy. The MED is widely used as a biological dose unit of exposure in clinical and experimental photodermatology and is based on a single acute UVR exposure. The MED is also used in the determination of the sun protection factor (SPF) of a sunscreen.

Whilst the MED is convenient indicator of individual sensitivity to UVR, the classification of people into skin phototype, as shown in Table 1.1, is a useful way of defining population acute and chronic sensitivity to UVR, and has been valuable in skin cancer epidemiology. An acute erythemal exposure, whether caused by solar simulating radiation (SSR) or UVA, is also associated with significantly increased sensitivity to mechanical and thermal stimuli [9]. In general, the higher the skin phototype the higher the MED but it must be stressed that neither skin phototype nor MED is predictive of each other on an individual basis because there is a considerable overlap of MED within skin phototypes [10]. Furthermore, it must be stressed that difference between the mean MEDs of sun-tolerant and sun-sensitive white skin phototypes is relatively modest. Typically, the MED of phototypes III/IV is about twice that of phototypes I/II [10,11]. Erythema from an acute erythemal exposure (5 MED) has been reported as being more persistent in skin phototype I compared with IV [12] and this may be related, in some as yet unknown way, to skin cancer susceptibility [13].

By definition, a single sub-erythemal exposure is below the threshold of detection by the eye. However, this does not mean that it has no effects because repeated daily sub-erythemal exposure results in clinically visible erythema after 2–3 exposures especially in sun-sensitive skin phototypes I and II [14,15], but less so with phototypes III/IV, even with dose as low as 0.25MED [14]. Thus, the MED is not a useful concept for the evaluation of erythema from repeated exposure that is best done with reflectance spectroscopy. This also highlights a limitation of the concept of SPF because a cumulative erythema can appear after a few days of correct sunscreen use that has resulted in daily sub-erythemal exposures.

Acute erythemal exposures of UVB and UVA result in a marked inflammatory infiltrate including neutrophils [16,17]. In one study an acute dose of 0.5 MED UVA, which did not result in any inflammatory infiltrate, was compared with repeated exposures of 0.5MED, 5 days/week for 6 weeks, which resulted in the presence of perivascular lymphocytes, some histiocytes and numerous mast cells [18].

COX dependent prostaglandin E₂ (PGE₂) is believed to be one of the mediators of UVR-induced erythema [19]. Epidermal COX-1 and COX-2 proteins are induced by an acute exposure of 3MED of a UVB-rich broad-spectrum source at a level that is comparable to ten consecutive exposures of 0.7MED, even though no erythema was reported with this protocol [20].

Overall, there is evidence that the effects of repeated sub-erythemal UVR on the skin are cumulative and that the regulation of erythema (inflammation) is better controlled in skin phototypes III/IV compared with I/II.

Epidermal melanin composition is a variable mixture of lighter alkali soluble, sulphur-containing phaeomelanin and darker insoluble eumelanin; skin phototypes that tan well have more eumelanin. Acute exposure to UVA induces a rapid but transitory grayish color known as immediate pigment darkening (IPD) which probably results from the photo-oxidation of existing melanins and the redistribution of melanocytic melanosomes from a perinuclear position into the peripheral dendrites [21]. IPD progresses into persistent pigment darkening (PPD), which may remain stable for up to 2 h post exposure. The biological functions of IPD and PPD are not known.

Delayed tanning, also known as melanogenesis, is primarily a response to UVB that results from increased activity and numbers of melanocytes. Tanning capacity is related to skin phototype as shown in Table 1.1. The action spectra for IPD and melanogenesis are completely different with the former showing a broad 340 nm peak in the UVA region [22] and the latter, similar to the erythema action spectrum, showing peak activity in the UVB region which is about three orders of magnitude more effective than UVA [7]. Melanogenesis becomes visible 3–4 days after an acute UVR exposure and is maximal from 10 days to 3–4 weeks depending on skin type and UVR dose. Melanocyte tyrosinase (the rate limiting enzyme for melanin synthesis) activity also increases, melanocyte dendrites elongate and branch, and melanosomes increase in number and size. Repeated acute sub-erythemal exposure doses also results in a gradual increase of tanning [23–25] with UVB containing sources including SSR. However, repeated exposure to sub-erythemal UVA [26,27] also induce delayed tanning. Melanogenesis is a multi-factorial process but there is a considerable body of evidence that DNA photodamage, and its repair, is a trigger. In other words, tanning may be a response to acute and persistent DNA photodamage [28].

The incidence of skin cancer is associated with skin phototype (see Table 1.1) and it is often stated that this is related to photoprotection by melanin. However, the skin phototype relationship with MED holds on vitiligenous skin through types II–VI indicating that factors other than melanin affect acute sensitivity to UVR [11].

UVR stimulates epidermal proliferation that results in stratum corneum (SC) thickening. An acute exposure of 1 MED SSR did not have any effect on Ki-67 expression, which is marker of epidermal proliferation, or epidermal thickening [29]. Repeated sub-erythemal SSR exposure, for approximately 2 weeks, of normal skin does result in modest SC thickening, and this is independent of skin type [30] and also results in the expression of Ki-67, even with doses as low as 0.25MED, though this dose did not result in epidermal thickening [29]. One analysis based on the relationship between MED and SC thickness indicated that the latter was an important factor in photoprotection [31] but this was not confirmed in a study in which volunteers were repeatedly exposed to sub-erythemal doses of SSR [24]. One study compared the effects of up to 24 weeks (twice weekly) of erythemal exposure (1 MED) on SC thickening of previously unexposed skin with SSR and UVA and with sub-erythemal UVA equivalent to the UVA content of the SSR source [26]. All spectra resulted in a time dependent increase of SC thickening that persisted for at least 12 weeks after the last exposure. Surprisingly, UVA including sub-erythemal UVA was more effective than SSR. There is evidence that the SC is a source of UVA-induced oxidative stress from acute exposure and that repeated sub-erythemal SSR exposure results in a decrease of SC catalase and an increase of protein oxidation, which may compromise its barrier function. Protein oxidation was not observed in the epidermis, possibly because of the epidermal induction of methionine-S-sulphoxide reductase (MSRA) [32] which is able to repair protein damage by ROS.

UVR absorption by epidermal DNA results in the formation of dipyrimidine lesions such as cyclobutane pyrimidine dimers (CPD) and pyrimidine (6–4) pyrimidone photoproducts (6–4 PP) [5]. Solar range UVR action spectroscopy shows that UVB is orders of magnitude more effective that UVA, as would be expected from the absorption spectrum of DNA [6]. The CPD, of which the thymine dimer (T = T) is the most common type, has been identified as critical photolesion for many of the acute and chronic effects of UVR. For example, the CPD is thought to initiate erythema [6], immunomodulatory cytokine release [33,34]. immunosuppression [35], and is known to be an important lesion in the generation of highly characteristic namely C→T or CC→TT transition mutations [36] in p53 that are common in actinic keratoses, BCC and SCC. BCC have also been found to harbour such mutations in the human homologue of the Drosophila “patched” Ptch gene [37], which suggests that this gene is important for this type of tumor. Its function is less clear than that of p53 but it is part of the hedgehog signal transduction pathway that transmits extracellular growth and differentiation signals to the nucleus. BCC are generally tumors of the elderly but the Gorlin-Goltz syndrome, a genodermatosis with autosomal dominant inheritance, predisposes to BCC at a very early age. This syndrome has been shown to be due to a mutation of the Ptch gene and Ptch^+/− mice develop BCC like tumors on exposure to broad spectrum UVR [38]. Interestingly, Ptch does not seem to play a role in human SCC leaving p53 the only gene known to lead to SCC upon inactivation.

The T = T has been identified in human epidermis in vivo in several studies and shows a UVR dose dependence in acute studies. These lesions are readily detected with acute sub-erythemal exposure in keratinocytes and melanocytes in human skin in vivo [39]. CPD, and other types of DNA damage, initiate one of two major pathways that are regulated by the p53 protein that is induced in the epidermis after acute erythemal UVR and repeated sub-erythemal exposure [29]. The first is nucleotide excision repair (NER) that can restore the integrity of the DNA, and is known to play a crucial role in the prevention of skin cancer. The most powerful human evidence for the major significance of DNA repair comes from the very high incidence of all types of skin cancer in xeroderma pigmentosum (XP) patients who are, to varying degrees, genetically deficient in NER [40]. There is evidence that DNA repair capacity; assessed in lymphocytes, plays a role in skin cancer in the normal population and that this is related to skin phototype. However, the relationship between skin phototype and DNA repair capacity has not been formally explored [41,42]. The NER of CPD is a relatively slow process such that many lesions are still present 24 hours post-exposure: in contrast the repair of the 6-4 PP is rapid [5,43]. The second pathway is a molecular cascade that results in the apoptotic death of the damaged cell [44]. UVR induced apoptotic cells are known as sunburn cells (SBC) that are UVR dose dependent with maximal expression at about 24 hours post-irradiation. In general, a significant increase of SBC is not seen unless the dose is approximately erythemal [29]. Repeated daily exposure of sub-erythemal SSR that results in the accumulation of CPD (Figure 1.1a) and p53 (Figure 1.1b) does not result in the accumulation of SBC [45]. The lack of SBC may be a means of protecting the integrity of the epidermis, at the expense of enhancing skin cancer risk. p53 expression is also seen on chronically sun-exposed skin [46,47] and has been proposed as a biomarker for skin cancer.

The presence of conical clonal patches of p53-mutated keratinocytes in normal human skin provides evidence that these mutations arise from mutated cells rather than from random mutation events. These p53 mutations in these clones confer resistance to apoptosis that allows keratinocytes to accumulate further UVR-induced mutations that may lead to skin cancer [48]. p53 clonal expansion is a function of chronic UVR exposure rather than time, because these clones regress in mouse skin in the absence of continuing exposure. Such clones are more frequent and larger in chronically sun-exposed skin but will not necessarily become a cancer. However, larger colonies provide bigger targets with a greater chance for secondary mutations that can lead to cancer. It has been suggested that p53 clonal expansion occurs by quantized colonization by the non-aggressive expansion of clones into compartments left empty by adjacent cells that have undergone UVR-induced apoptosis. Overall, our current understanding of non-melanoma skin cancer is that DNA photodamage is the initial photomolecular event that triggers a chain of cellular, mutational [36], and immunological events [49] that may lead to a skin tumor.

There is emerging evidence that CPDs, but not (6–4) photoproducts, can be formed by energy transfer reactions originating from UVA chromophores [50], which means that some of the C→T or CC→TT transitions found in skin cancers, normally associated with UVB, may be attributable to UVA. In vitro and in vivo studies on human skin have demonstrated UVA-induced oxidative DNA lesions such as 8-oxo-7,8-dihydro-2'-deoxyguanosine which have been reported under laboratory conditions [51], and 8-hydoxy-2'-deoxyguanine [52] and UVA signature mutations (AT→CG transversions) have been reported in p53 in AK, BCC and SCC. The chromophores for these lesions are not known.

The molecular relationship between UVR and MM is much less well understood. Studies on familial MM have identified an important role for the CDKN2A gene that codes for two distinct tumor suppressor proteins: p16^INK4a and p14^ARF which affect the retinoblastoma (Rb) and p53 pathways, respectively [53]. There is emerging evidence from in vitro animal and human studies of significant interaction between UVR and the p16^INK4a/Rb pathway, including limited evidence of UVR signature mutations in human MM [53,54].

Mouse studies have enabled the determination of an action spectrum for SCC that is shown in Figure 1.2, along with the standard human action spectrum for erythema. Comparison of these data shows a high degree of similarity, which provides indirect evidence for a shared chromophore, which is most probably DNA, though there are probably as yet undefined chromophores in the UVA region. There has been a considerable amount of debate on the action spectrum for MM, with some studies suggesting a major role for UVA. However, recent studies on a new transgenic mouse (HGF/SF) model for MM indicate that UVB is much more important than UVA [55].

It is well established that UVR affects cell-mediated acquired immunity and it is thought that this plays a significant role in photocarcinogenesis [49] (and chapter by Halliday in this volume). Langerhans cells (LH), the professional antigen presenting cells (APC) of the epidermis, are sensitive to UVR which results in their depletion and changes in their morphology such as loss of dendricity. In general, an approximately acute erythemal dose is necessary to result in significant LC depletion [15], but there is evidence that repeated daily sub-erythemal exposure has a cumulative effect that also results in loss of LC density [15,29,45,56], which is observed in chronically sun-exposed skin [58,59]. This leads to the conclusion that both acute erythemal and chronic sub-erythemal UVR exposure results in LC loss and that this is likely to influence antigen presentation. However, as will be evident from the paragraph below, acute sub-erythemal exposure, which may have no obvious effect on LC density, can affect antigen presentation. Exposure to erythemal SSR also results in the epidermal appearance of CD11b+ macrophages [15] that result in altered antigen presentation that is also observed with repeated sub-erythemal exposure.

Acute erythemal exposure (typically 3MED) has been shown to induce a wide range of immunomodulatory cytokines (e.g. TNFα, IL-10) in human skin in vivo measured at the mRNA [60] and protein [61,62] levels, but the effects of acute sub-erythemal exposure on cytokine profile in human skin does not seem to have been investigated. One study has assessed the effects of ten consecutive sub-erythemal (0.7MED) exposures with a UVB rich source. This resulted in an increase of whole skin mRNA for IL-1β, IL-6, IL-10 and TNFα, but this increase was smaller than that observed for a single 3MED exposure [20].

Most of the studies of UVR on human immune function have been with the contact hypersensitivity (CHS) or delayed type hypersensitivity (DTH) models in which the initial contact with the antigen is delivered, respectively, via a topical or sub-cutaneous route. The effect of UVR can be assessed on either the sensitization or the elicitation arms of the model. Several studies have shown that a single acute sub-erythemal exposure of SSR can suppress the sensitization arm of the CHS response. One study, comparing phototypes I/II with III/IV, has shown that an acute erythemal exposure is necessary to suppress this response in skin phototypes III/IV in contrast to the immunosuppressive effects of 0.25 to 0.5MED in phototypes I/II [63]. Such data show that erythema is a poor indicator of the immunosuppressive effects of UVR. Repeated daily sub-erythemal (0.3MED) fluorescent SSR, for up to 30 days, has been shown to have a cumulative effect on suppression of the sensitization phase of the CHS response [64], which indicates a lack of a photoadaptive response. In general, repeated erythemal exposure is necessary to suppress the elicitation arms of the CHS and DTH responses [65]. The use of tanning devices can also affect cutaneous immunity. One study showed that 10 full-body sub-erythemal exposures over a two-week period suppressed the induction and the elicitation arms of the CHS response [66].

The wavelength dependence for the immunosuppressive effects of UVR is still poorly defined, especially in humans [67]. There is no doubt that UVB is immunosuppressive but the role of UVA, and its possible interaction with UVB, requires further investigation; to date, much of the evidence for a role for UVA has been indirectly obtained from sunscreen studies.

Studies on the acute effects of UVR on skin may give misleading results because they do not take any adaptive responses into account, the most obvious of which is tanning. The possible photoprotective role of a tan has been assessed epidemiologically in MM with one study suggesting a benefit [68] that was not confirmed in another study [69]. Tanning has been shown to offer some modest photoprotection against DNA photodamage and erythema from a subsequent challenge dose [28], but it has to be remembered that DNA photodamage is acquired during the tanning process [20,25,70], and that the acquisition of such damage is likely to contribute to mutation and skin cancer. A review of the literature suggests that tanning offers a modest protection factor in the region of 2–3 [28]. There is evidence that adaptation is skin phototype dependent. Repeated SSR exposure, with a given sub-erythemal dose, in skin phototypes I/II results in an accumulation of erythema that is not seen with phototype III/IV [24,25], despite the higher physical UVR doses accumulated by the latter, because of their higher MEDs. This indicates that sun-tolerant phototypes show better regulation and control of the inflammatory response than sun-sensitive phototypes. Furthermore, there is some evidence of better, possibly inducible, repair of CPDs in sun-tolerant phototypes ([25,71] and chapter by Beer and Hearing in this volume). Photoadaptation to the immunological effects of UVR has been studied in pooled groups of phototypes II/III. One study, of up to 30 consecutive exposures of ∼0.3MED fluorescent SSR, showed increasing suppression of the induction arm of the CHS response, which indicates a cumulative response lacking in adaptation [64]. A subsequent similar study by the same group assessed the effect of ten prior consecutive 0.7MED exposures of a UVB-rich source on the immunosuppressive effects of a challenge dose of 3MED [20]. A comparison with a 3MED exposure alone showed no evidence of adaptation. This study also looked at photoadaptation to other endpoints including cytokine mRNA expression, COX-1 and 2 proteins and CPD. In general, there was little if any evidence of protection from damage by the prior exposures.

Photoaging is distinct from normal chronological aging. Clinically, photoaged skin is dry, deeply wrinkled, inelastic, leathery and telangiectatic, often with irregular pigmentation, freckling and lentigo formation. Histologically, such skin shows marked quantitative and qualitative abnormalities, particularly of the dermal connective tissue, including the accumulation of abnormal elastotic material, namely, elastosis, and proteoglycans. Furthermore, the degradation and disorganization of collagen fibrils, responsible for the strength and resilience of skin, have been observed.

There are no comprehensive epidemiologic data on photoaging but sunlight is firmly implicated through comparison of adjacent sun-exposed and sun-protected skin, and animal studies provide conclusive evidence for a significant role for UVR, especially UVB for elastosis [72] which has an action spectrum similar to that for human erythema. A few studies have assessed the role of repeated exposure on dermal connective tissue. One study compared the effects of twice weekly doses of 1 MED SSR, 1 MED UVA as well as the UVA content of 1 MED SSR (i.e. low dose UVA) for up to 24 weeks [26]. SSR resulted in an increase of elastic tissue content whereas, both UVA protocols resulted in a reduction that persisted for at least 12 weeks post-irradiation. These data suggest that chronic exposure to sub-erythemal UVA plays an important role in photoageing. Another study compared repeated sub-erythemal (0.5 MED) exposure for SSR and UVA given 5 days per week for 6 weeks. In comparison with acute exposures of up to 2MED (which showed no difference from non-irradiated controls) SSR and UVA resulted in an increase in lysozyme deposition on elastin fibers (a marker of UVR damage of this tissue), although the effect of SSR was not significant [18]. More recent studies have shown that repeated sub-erythemal exposure (nine daily exposures) with SSR, or a spectrum that simulates sunlight at temperate latitudes, resulted in a loss of glycosoaminoglycans (GAGs) (with doses as low as 0.25MED) and collagen alterations associated with photoaging [29].

Overall, these studies show that repeated UVR exposure, including sub-erythemal UVR exposure results in dermal damage after a relatively short period of time. It is reasonable to suppose that repeated exposure, over decades, will result in changes that manifest as clinical photoaging.

It is thought that UVR-induced tissue-degrading matrix metalloproteinases (MMPs) may be involved in photoaging [73]. MMPs belong to a family of zinc-dependent endopeptidases that degrade structural proteins such as collagens and elastin in connective tissue. Their proteolytic activity is regulated by endogenous tissue inhibitors of metalloproteinases (TIMPs) that inhibit the active MMPs by the formation of tight noncovalent 1:1 complexes. Single exposures of UVB [74] and environmentally relevant doses of SSR [75] induce MMPs and TIMPs gene expression in human skin within 24 h, but the induction of MMP-1 is much greater than TIMP-1 at the mRNA level. In the case of SSR, an erythemal exposure is necessary [75], but the induction of protein MMP-1 expression has been reported with single very low doses (<0.1 MED of UVB with probable UVC contamination) [74]. MMP induction occurs through the activation of the AP-1 transcription factor. Four exposures of 0.5MED UVB (filtered to remove wavelengths <290 nm) given at 48 hour intervals resulted in sustained MMP activity, that was not different from a single exposure [76]. A current model of photoaging proposes that UVR-induced MMPs degrade the dermal matrix, which is followed by imperfect matrix repair. With repeated MMP induction over time, the faulty repair gives rise to a “solar scar” that manifests as skin wrinkling [76]. The chromophores for photoaging have not been investigated; these may well be molecules within cell surface receptors, as well as DNA. There is also evidence that ROS may play a role.

Vitamin D₃ (cholecalciferol) is vital for bone health and its cutaneous photosynthesis is the only well-established benefit of UVR exposure. The plasma membrane bound chromophore 7-dehyrocholesterol (7-DCH or provitamin D₃) is photoconverted to previtamin D₃ that undergoes a slow thermal isomerization to Vitamin D₃. Action spectra data indicate that UVB (∼295–315 nm) is responsible for Vitamin D₃ production [77], which is also the peak region for sunburn and non-melanoma skin cancer. Vitamin D₃ is biologically inert and requires two hydroxylation steps: the first in the liver to 25-hydroxyvitamin D₃ (25-(OH)D₃) and the second in the kidney to produce the biologically active form 1,25(OH)₂D₃ (calcitriol). The plasma concentration of 25-(OH)D₃ is easily measured and is the conventional indicator of Vitamin D₃ status. More recently, it has been established that many tissues and cells can synthesize and respond to calcitriol as they possess both 1α-hydroxylase and Vitamin D receptors. This has focused attention on the non-calciotropic effects of Vitamin D₃. There is emerging evidence, mostly epidemiologic (i.e. based on association with latitude), that low solar UVB exposure correlates with poor Vitamin D₃ status that is in turn associated with increased risk of internal malignancies (e.g. colorectal and prostate cancer) and autoimmune disorders [78,79]. This link between Vitamin D₃ status and non-calciotropic health benefits remains highly controversial [79], however, and in any case it could be argued that Vitamin D₃ status is best maintained by improved diet and supplementation, as achieved in several controlled studies [80].

There is no standard definition of optimal Vitamin D₃ status but it is generally accepted that plasma 25-(OH)D₃ below 25 nmol/l constitutes deficiency as it is strongly associated with rickets and osteomalacia. It has been suggested that a level of 50–80 nmol/l is required to optimize bone health, particularly in the elderly [81]. A new and controversial category of Vitamin D “insufficiency” has been defined as higher Vitamin D levels than those causing rickets or elevated parathyroid hormone levels but below levels that are statistically associated with other possible Vitamin D benefits in population studies [80].

In temperate regions above approximately 45° latitude there is insufficient ambient UVB during September/October to March/April to synthesize Vitamin D₃ [82]. Therefore, body stores, dietary intake, supplementation and UVB tanning (but not UVA) devices are the only sources during this time. Several studies have shown substantial seasonal fluctuation of 25-(OH)D₃ with significant proportions of populations being Vitamin D₃ insufficient (see definition above) in late winter and early spring [83]. People who habitually avoid sunlight during the summer, the elderly and populations with high constitutive pigmentation have been shown to have relatively low levels of 25-(OH)D₃ throughout the year [84,85].

It is not known how much UVB and dietary Vitamin D are needed to maintain plasma 25-(OH)D₃ > 50 nmol/l. It is widely advocated that 15 minutes’ exposure of the face, arms and hands to noonday summer sunlight two or three times a week is “sufficient” for Vitamin D synthesis, but it is unclear if this will be enough to achieve adequate Vitamin D status at the end of the summer. In any case, the validity of this statement has never been formally tested. It is an estimate based on human studies [77,86–88] using artificial sources of UVR that have serious inadequacies, such as low subject numbers (n = 2 – 8), the use of single whole body exposures with UVR sources with a spectrum completely different to sunlight and the measurement of endpoints other than 25-(OH)D₃.

Sunscreens are designed to prevent erythema from acute solar exposure and their efficacy is indicated by their sun protection factor (SPF) that is not necessarily indicative of protection from other types of acute photodamage, let alone chronic photodamage. Indeed the evidence for a long-term benefit of sunscreen use is somewhat limited with some evidence for protection from SCC, but none for BCC and MM [89–91].

The relationship between protection against erythema and non-erythema acute endpoints is very important in long-term photoprotection for which sunscreen use is widely advocated. For example, a sunscreen that protected against erythema but gave a much lower level of protection against immunosuppression could give a false sense of security and actually increase skin cancer risk, because more immunosuppressive damage might be done for a given level of erythema with the sunscreen that the same level of erythema without the sunscreen.

Reassuringly, one study has indicated that the SPF is equivalent to a “DNA protection factor” [92] which is to be expected if DNA is a major chromophore for erythema. Another study has confirmed the ability of a sunscreen (SPF = 15) to protect against CPD after four daily exposures of 2 MED SSR [93]. During recent years there has been considerable interest in the ability of sunscreens to prevent acute UVR-induced immunosuppression and the development of an understanding of the relationship between the SPF and immune protection factor (IPF); it is clear that these protection factors (both of which are ratios) are not necessarily equivalent which would be expected if the photoprotection endpoints shared a common single chromophore. The nature of this relationship is still poorly understood, but there is evidence that immunoprotection is more dependent on attenuation of UVA than protection from erythema.

There are very few data on the role of sunscreens in the prevention of photodamage, including erythema, from repeated doses of UVR. One study has shown that a broad-spectrum sunscreen (SPF = 8) could inhibit the photoageing related changes observed after 6 weeks (5 days/week) exposure to 1 MED SSR [94]. Another study has shown that a broad-spectrum sunscreen with an SPF of 7 was able to protect against the effects of 11 consecutive days of sub-erythemal SSR exposure; these included the accumulation of erythema, CPD, p53 protein and the loss of LC [45]. A field study in Australia has shown that sunscreen use protects against p53 accumulation on the back of the hand [46].

The relationship between acute and chronic photodamage is crucial to the understanding of the possible long-term benefits of sunscreen use. Prospective sunscreen studies are very difficult to carry out, whereas short-term repeated exposure studies with relevant biomarkers are possible and may well be valuable in assessing important non-biological factors such as behavior and compliance.

The basis of skin phototype is not understood but it is generally, though not individually, predictive of acute UVR sensitivity, tanning capacity and skin cancer risk. Photoprotection by melanin almost certainly plays a role, but other factors such as possible differences in DNA repair capacity and susceptibility to photoimmunosuppression may play a role and require further investigation. Considerable molecular and cellular damage to the skin can be done with acute sub-erythemal exposure. At present there is a large gap in our understanding of how the acute effects of UVR, especially with repeated sub-erythemal exposure, impact on the well-established chronic effects such as skin cancer and photoageing. In a large part this is simply due to a lack of data on the molecular and cellular effects of repeated sub-erythemal exposure, which would provide information of the adaptive responses of the skin, or help identify biomarkers (e.g. p53 accumulation) for outcomes such as skin cancer. The acquisition of such data would not of course provide a direct link with effects that take decades for clinical manifestation. However, they would certainly be more useful than the study of single acute erythemal exposure. Such data would also be better for the development of photoprotective strategies for sun-sensitive phototypes. The accumulation of damage with sub-erythemal exposure suggests that the oft-given advice that chronic damage can be avoided by avoiding sunburn may be misleading as a public health message.