Does an active sun exposure habit lower the risk of venous thrombotic events? a d-lightful hypothesis

Journal of Thrombosis and Haemostasis, 7: 605–610 Does an active sun exposure habit lower the risk of venousthrombotic events? A D-lightful hypothesis P . G . L I N D Q V I S T * , E . E P S T E I N   and H . O L S S O N à*Department of Obstetrics and Gynecology, Clintec, Karolinska University Hospital, Huddinge, Stockholm;  Department of Gynaecology, Lund University Hospital, Lund University, Lund; and àDepartment of Oncology and Cancer Epidemiology, Lund University Hospital, Lund University, To cite this article: Lindqvist PG, Epstein E, Olsson H. Does an active sun exposure habit lower the risk of venous thrombotic events? A D-lightful hypothesis. J Thromb Haemost 2009; 7: 605–10.
Summary. Background: Venous and arterial thrombotic com- Venous thromboembolism (VTE) events constitute a major plications exhibit a seasonal variation, with risk peaking in cause of female morbidity and mortality. The risk of VTE winter and dropping to a nadir in summer. We sought a possible increases with advancing age, the presence of inherited or correlation between sun exposure habits and venous thrombo- acquired thrombophilias, hypofibrinolysis, surgery, hormonal embolism (VTE) events. Methods: This was a cohort study use [combined oral contraceptives (COCs) or hormone replace- comprising 40 000 women (1000 per year of age from 25 to ment therapy (HRT)], immobilization, overweight, pregnancy, 64 years) who were drawn from the southern Swedish popu- and malignancy [1,2]. There are several studies indicating that lation registry for 1990 and followed for a mean of 11 years.
the risk of VTE is greater in the winter months than in the Seventy-four per cent answered an inquiry at the inception summer months [3–6]. Coronary heart disease (CHD) and of the study (n = 29 518), and provided detailed information other arterial thrombotic complications have shown a similar on their sun exposure habits. Cox regression analysis was seasonal pattern [7–10]. No plausible explanation has yet been used with the presence of VTE as a dependent variable and given for the seasonal variations in thrombotic complications.
selected demographics as independent variables. The main Vitamin D levels have been demonstrated to have a similar outcome was the relationship between VTE and sun exposure seasonal variation, with a nadir occurring in winter [11].
habits. Results: Swedish women who sunbathed during the Humans obtain vitamin D from exposure to sunlight, diet, or summer, on winter vacations, or when abroad, or used a dietary supplements [12]. Most dietary products are low in tanning bed, were at 30% lower risk of VTE than those who vitamin D. Therefore, the major source of vitamin D is did not. Risk estimates did not change substantially after ultraviolet B (UVB) radiation (wavelength between 290 and adjustment for demographic variables. The risk of VTE 315 nm), which penetrates the skin and converts 7-dehydro- increased by 50% in winter as compared to the other seasons; cholesterol to 25-hydroxycholecalciferol vitamin D3 (25-OH- the lowest risk was found in the summer. Conclusions: Women with more active sun exposure habits were at a significantly OHVitD into its active form, 1a,25(OH)2 vitamin D3 lower risk of VTE. We speculate that greater ultraviolet B (1,25VitD), takes place mainly in the kidney [12].
light exposure improves a personÕs vitamin D status, which In our study, the Melanoma Inquiry of Southern Sweden in turn enhances anticoagulant properties and enhances the (MISS), we followed 40 000 women prospectively for a mean period of 11 years, obtaining detailed information on their sunexposure habits, as well as such established risk factors for VTE Keywords: sun exposure habits, venous thromboembolism.
as age, malignancy, number of births, and hormonal treatment.
This longitudinal cohort study was carried out in order to assesshow womenÕs sun exposure habits influence their risk of VTE.
Correspondence: Pelle G. Lindqvist, Department of Obstetrics and The study was approved by the Ethics Committee of Lund Gynaecology, Clintec, Karolinska University Hospital, Huddinge, University (LU 632-03). In 1990 the MISS study was initiated.
Kvinnokliniken K 57, SE-14186 Stockholm, Sweden.
One thousand native-born Swedish women per year of age, Tel.: +46 708 992545 or +46 40 158910.
from 25 to 64 years (n = 40 000), with no history of malig- nancy, were chosen from the general population registry of the Received 26 October 2008, accepted 25 January 2009 South Swedish Health Care Region by computerized random Ó 2009 International Society on Thrombosis and Haemostasis selection. Twenty-seven women could not be contacted, leaving Hormone use at the inception of the study, that is, use of 39 973 as the study cohort, representing 20% of the south COCs or HRT, was introduced as a dichotomized variable Swedish female population in the selected age groups. The [never use (reference), ever use]. The question posed was: have women were invited to complete a standardized written you used/are you using combined oral contraceptives? Women questionnaire concerning risk factors for malignant melanoma.
not answering the question were considered never to have used The initial inquiry was made between 1990 and 1992, and a written follow-up was conducted between 2000 and 2002. The Smoking habits were also recorded at the inception of the questionnaire inquired into several items of potential interest study. They were categorized into the following subgroups: for thrombosis risk, such as number of births, marital status, non-smokers (reference); those who had smoked fewer than and educational level, and included detailed questions regard- 100 000 cigarettes in their lifetime; and those who had smoked ing sun exposure habits. The sun exposure questions at the 100 000 cigarettes or more (based on how participants had inception of the study were: (i) how often do you sunbathe characterized their cigarette smoking in mean consumption at during the summer? (never, 1–14 times, 15–30 times, > 30 times); (ii) do you sunbathe during the winter, such as during Drinking habits were noted at the time of the initial vacations to the mountains or the Alps? (never, 1–3 days, 4– questionnaire by quantity of beer, wine and spirits consumed 10 days, > 10 days); (iii) do you use a sun bed? (never, 1–3 per month. Alcohol intake was categorized into five subgroups times, 4–10 times, > 10 times per year); (iv) do you work by equivalent amounts of alcohol consumed: no consumption outdoors during the summer (no, yes); and (v) do you go (reference); < 5 g day)1; 5 to < 10 g day)1; 10–15 g day)1 abroad on vacation to swim and sunbathe? (never, once every (moderate consumption); and > 15 g day)1.
year or two, once a year, two or more times a year). For Weight and height were recorded at the second interview, analysis, the five questions were dichotomized into negatives and body mass index (BMI) was calculated as kg m)2. BMI (no/never) and affirmatives (varying positive frequencies). All was classified into three groups: < 25 (reference); 25 to < 30 data regarding risk factors, apart from cancer diagnosed during the study period, were collected from the initial written inquiry.
The level of regular exercise was estimated at the second At the follow-up interview, women were asked about long-term interview by answers to the question – ÔIn addition to your medications that they may have taken and the presence of other usual work, do you exercise regularly?Õ: No, Do you go for a diseases (including VTE). In order to determine whether there walk once a week? Do you go for a walks several times a week? was a dose–response relationship, a new categorized three-part Do you bicycle, swim, participate in gymnastics, dancing, or dummy variable was created (no/never, sometimes, or more similar activities one or more times a week (i.e. strenuous exercise)? Physical exercise was then divided into three catego- The unique personal identification number assigned to each ries: none, take walks one or more times a week, or strenuous Swedish resident allowed us to ascertain all deaths and causes of death from the National Cause of Death (NCD) register.
The incidence of VTE was determined both through responses to the follow-up questions regarding disease and long-termmedication, and by means of entries in the National Patient Analysis of characteristics of selected variables were performed Registry (NPR), which records all women who have been with Cox regression analysis using 95% confidence intervals.
The presence of VTE was used as a dependent variable, and Those women who were invited to participate in the MISS Ôtime-at-riskÕ as a time variable. Time-at-risk was defined as study were sought by ICD 9 code numbers for VTE – 634G or time from initial participation in the study to VTE, death, or 31 634H, 635G or 635H, 636G or 636H, 637G or 637H, 638G or December 2002, whichever came first. As increasing age is a 638H, 639G or 639H, 671D, 671E, 671F, 673C, 451B, 452, 325, strong risk factor for VTE, it was introduced as a categorized 437G, 572B, 453 (C, D, W, or X) or 415B – or the variable and included for adjustment in all risk estimates. All calculations were performed using SPSS software (Statistical O873, O879, O225, O229, O882, I802, I803, I81*, I82*, I636, Package for the Social Sciences, SPSS Inc., Chicago IL, USA), I676, K550, or I26*. The personal identification number and and P-values < 0.05 were considered to be statistically the above-mentioned diagnosis codes were cross-matched in the NCD and NPR. Thus, the registered cases, and dates, ofVTE events were established for all women in the cohort, that is, both those included and those not included in the analysis ofrisk factors. From the register, we collected all VTE events up In Table 1, we present the characteristics of the women in the to 31 December 2002. Malignancy was defined as the diagnosis study cohort. There was an increased incidence of VTE among of any type of malignancy during the study period prior to 31 women with less than 9 years of schooling, unmarried women, December 2002. Information was gathered from both Regional and widows. As compared with those who had given birth on and National Cancer Registries. Vital statistics were deter- one or two occasions, nulliparous and multiparous women were at increased risk. Women who were diagnosed with cancer Ó 2009 International Society on Thrombosis and Haemostasis Sun exposure and venous thrombotic events 607 Table 1 Demographic characteristics of women with and without venous during the study period were at four-fold increased risk of thromboembolism (VTE) from inquiry at inclusion VTE. At the initial interview, data were gathered from 29 518 women of the 39 973 in the total cohort (74%), representing 317 290 woman-years; 24 098 women answered the follow-up Table 2 shows the age-adjusted analysis of sun exposure habits and VTE risk. Those who used a sun bed, sunbathed during winter vacations or during the summer, or who sunbathed abroad, were all at about 30% reduced risk of VTE. There were only minor changes in risk estimates when adjusting for demographic characteristics (model 1). In mod- el 2, an adjustment for smoking habits and alcohol consump- tion was added, and in model 3, exercise and BMI were added by means of data obtained from those answering the follow-up inquiry (i.e. retrospective information). The relative risks (RR) for smoking habits, moderate alcohol consumption, obesity and strenuous exercise were 1.4, 0.4, 2.5, and 0.5, respectively.
All differed significantly from the reference groups. A dose– response relationship was not found between sun exposure habits and risk of VTE. The risk among those sunbathing 1–14 times during summer was almost identical to that for those sunbathing ‡ 15 times. Women using sun beds one to three times during the course of a year were at similar risk as those using them at least four times per year. The risk to those who frequently sunbathed abroad or during winter vacations was not significantly lower than the risk to those who did not.
Figure 1 shows the mean hours of sunlight per month by season and the number of VTE events annually for the whole study population (n = 39 973). It is notable that the risk ofVTE in the winter is between 40% and 60% greater than in CI, confidence interval; RR, relative risk. All information, except that other seasons. As compared to winter, the risk is lower in spring for cancer (follow-up data), was gathered at study inclusion andanalyzed with Cox regression analysis with age adjustment.
[30% lower, odds ratio (OR) = 0.7], in summer (OR = 0.6), *Some women did not answer all questions.
Table 2 Sunbathing habits and risk of venous thromboembolism (VTE); bivariate age-adjusted analysis CI, confidence interval; RR, relative risk. Data were analyzed with Cox regression analysis with age adjustment. Model 1: Adjusted for age,education, marital status, number of births, and cancer during study period. Model 2: Adjusted for age, education, marital status, number of births,cancer during study period, alcohol consumption, and smoking habits. Model 3: Adjusted for age, education, marital status, number of births,cancer during study period, body mass index, and physical activity, including those who answered the follow-up inquiry.
Ó 2009 International Society on Thrombosis and Haemostasis in vivo. PAI-1 is a marker of fibrinolytic activity, and the level is positively associated with cardiovascular risk. The t-PA Ag level is useful as a marker of endothelial dysfunction. Endo- thelial cells show 1,25VitD receptor activity together with a1a-hydroxylase enzyme for local 1,25VitD production from25-OHVitD. In patients with type 2 diabetes mellitus with low 25-OHVitD status, supplementation with vitamin D(vitamin D2, 100 000 units, single dose) resulted in decreased blood pressure and improved flow-mediated vasodilatation (i.e.
improved endothelial function) [16]. There are several possiblemechanisms by which vitamin D may improve endothelial function: indirectly, by reducing blood pressure [16,17]; bydecreasing vascular resistance [16]; by acting as an immuno- modulator, preventing excessive expression of inflammatory cytokines (tumor necrosis factor-a, interleukin-6) [18,19]; and Fig. 1. Number of women with venous thromboembolism (VTE) and by increasing interleukin-10 expression [20]. Thus, there seems mean hours of sunlight per month by season.
to be a role for vitamin D in maintaining the integrity of thevascular endothelium.
Several factors affect the ability to form 25-OHVitD via the skin. Clothing and sunscreen are effective in preventing We found women with active sun exposure habits to be at vitamin D synthesis [21,22]. Sun block with sun protection significantly lower risk of VTE. This finding was constant after factor (SPF) 15 absorbs 99% of the incident UVB radiation, adjustment for demographic variables. It was also constant thus preventing most of the 25-OHVitD synthesis [23]. Melanin after adjustment for smoking habits, alcohol consumption, is extremely efficient in absorbing UVB radiation: increased BMI, and physical exercise. The results for these lifestyle skin pigmentation markedly reduces 25-OHVitD synthesis in variables were recently reported [13]. Furthermore, a seasonal the skin [19,22]. African Americans with very dark skin have an pattern was noted in the risk of VTE: it increased by about equivalent SPF of 15; that is, their ability to synthesize 50% in winter, when the hours of sunlight were few. We believe vitamin D is reduced by 99% [19]. A large cohort study in the that our finding of reduced risk of VTE with increased sun USA reported a 60% increased risk of VTE among blacks, as exposure might offer a plausible explanation for the seasonal compared with whites [24]. Thus, our hypothesis might be variation in the incidence of VTE [3–6] Women who exposed relevant in explaining some of the racial differences concerning themselves more often to the sun or to artificial UVB light VTE. Sunbathers using a sun bed have been shown to have presumably improved their vitamin D status [14]. As women robust 25-OHVitD levels [14]. In addition, the angle at which with more active sun exposure habits were at lower risk of the sun shines upon the earth has a major effect on the amount VTE, we speculate that variations in vitamin D status might be of UVB light that reaches the surface [19]. Therefore, not much a possible cause for the seasonal variations in thromboembolic vitamin D is synthesized in the morning or late afternoon.
complications. This study lends support to the previously Furthermore, the capacity to synthesize 25-OHVitD in the skin reported health benefits of sun exposure [12]. In the past, the declines with age (a 70-year-old person has about 25% of the effect of cold has been proposed as a cause of the seasonal variation in these risks [7]. With more hours of sunlight, there We have found no trials designed to test the ability of will be less cold weather and improved vitamin D status. Thus, vitamin D to prevent VTE. However, a randomized controlled our hypothesis is in agreement with the correlation between study of 250 patients with prostate cancer, half of whom were cold and thrombotic complications [7], but not with cold given 45 lg of calcitriol (1,25VitD) weekly, found an unex- pected significantly lower risk of thrombotic events (two vs. 11events) [25].
By what mechanism might sun exposure lower the risk ofVTE? The active metabolite of vitamin D is 1,25VitD. It has been Strengths of the present study are its use of an unselected large shown to have anticoagulant properties by upregulating cohort drawn from the national population registry and the thrombomodulin and downregulating tissue factor [15], Thus, fact that the information was obtained at the inception of the the decreased vitamin D levels during winter move the study. The combination of administrative data and the coagulation balance towards hypercoagulation, as compared questions put to women at follow-up about diseases and with summer. Levels of 25-OHVitD have been shown to be long-term medication is a strength; we are confident that most inversely related to plasminogen activator inhibitor-1 (PAI-1) outpatient VTE events have been recorded. However, we have and tissue-type plasminogen activator antigen (t-PA Ag) levels no data on the incidence of misclassification among the Ó 2009 International Society on Thrombosis and Haemostasis Sun exposure and venous thrombotic events 609 diagnosis numbers. We also lack data on familial thrombosis cardiovascular system, and the lower risk of malignant cancers or prior thromboembolic events. There is a possibility that women with prior VTE events travel less, and thus have a lower Future studies could be designed to differentiate between level of sunbathing. An additional weakness is that we lack the effects of cold and sunlight regarding the risk of information on other established risk factors, such as surgery, thrombotic complications. In a case–control set-up, it would injury, and immobility. The assumption was made that sun be possible to determine 25-OHVitD levels in cases and exposure habits did not change over time and, consequently, information from one assessment alone was used in the models.
We believe that our study offers novel epidemiologic data This is a common assumption in cohort studies, and it tends to together with an etiologic hypothesis regarding the cause of lead to underestimation of the risk. Another potential short- seasonal variations in the risk of thrombotic complications.
coming is that we lack data on what method was used for Our findings indicate that women who described themselves as diagnosing a VTE event. However, the employment of having more active sun exposure habits were at a significantly objective methods is widespread in Sweden, as the cost of lower risk of VTE than those who did not.
verifying VTE events is borne by the Swedish social securitysystem. Thus, we do not believe there is a substantial ÔoverdiagnosisÕ. Clinically unrecognized VTE events mighthave been missed (for example, a sudden death caused by This study was supported by the ALF (Faculty of Medicine, pulmonary embolism might have gone undetected without an Lund University, Region Ska˚ne), the Swedish Cancer Society, autopsy). However, in order to minimize this shortcoming, the and the Swedish Medical Research Council, and received study population was cross-matched against the cause of death funding from Lund University Hospital and Region Ska˚ne, the Gustav V Jubilee Fund, and the Kamprad Foundation.
Women with one or two prior deliveries are at lower risk of VTE than are nulliparous women. Those who have made it through one or two high-risk periods without a VTE event haveshown themselves not to be at high risk. The few who have had a The authors state that they have no conflict of interest.
prior VTE event are presumably treated with thromboprophy-laxis in high-risk situations. This may account for the low risk among those with one or two prior deliveries. Our finding ofincreased risk among nulliparous women is in agreement with 1 Lisman T, de Groot PG, Meijers JC, Rosendaal FR. Reduced plasma studies previously conducted in Sweden [26], but in partial fibrinolytic potential is a risk factor for venous thrombosis. Blood 2005;105: 1102–5.
opposition to the generally accepted view that a high number of 2 Dahlba¨ck B. Blood coagulation. Lancet 2000; 355: 1627–32.
prior births is a risk factor for VTE. A longitudinal cohort study 3 Gallerani M, Boari B, de Toma D, Salmi R, Manfredini R. Seasonal using data from inception is not the most appropriate design for variation in the occurrence of deep vein thrombosis. Med Sci Monit studying the relationship between hormonal use and the acute precipitation of a disease such as VTE, where the highest risk 4 Clauss R, Mayes J, Hilton P, Lawrenson R. The influence of weather and environment on pulmonary embolism: pollutants and fossil fuels.
occurs shortly after initiation of therapy. Thus, the estimates for Med Hypotheses 2005; 64: 1198–201.
ever-users of COCs at inception of the study should not be 5 Gallerani M, Boari B, Smolensky MH, Salmi R, Fabbri D, Contato E, regarded as valid estimates of present use: because of the age Manfredini R. Seasonal variation in occurrence of pulmonary embo- distribution in the present study, over 80% were prior users.
lism: analysis of the database of the Emilia-Romagna region, Italy.
Prior users are expected to be less likely to have a VTE event 6 Ho JD, Tsai CY, Liou SW, Tsai RJ, Lin HC. Seasonal variations in during the study period, for the same reason as above. In the occurrence of retinal vein occlusion: a five-year nationwide addition, a longitudinal cohort study with data from inception population-based study from Taiwan. Am J Ophthalmol 2008; 145: will tend to underestimate the risk of current use. A nested case– control design would have been more appropriate.
7 Nayha S. Cold and the risk of cardiovascular diseases. A review. Int J The SunSmart program promoted by Cancer Research UK Circumpolar Health 2002; 61: 373–80.
8 Ricci S, Celani MG, Vitali R, La Rosa F, Righetti E, Duca E. Diurnal (spend time in the shade between 11 a.m. and 3 p.m., make and seasonal variations in the occurrence of stroke: a community- sure that you never burn, aim to cover up with a T-shirt, hat, based study. Neuroepidemiology 1992; 11: 59–64.
and sunglasses, remember to take extra care with children, and 9 Khan FA, Engstrom G, Jerntorp I, Pessah-Rasmussen H, Janzon L.
then use factor 15+ sunscreen) has been subject to revision; Seasonal patterns of incidence and case fatality of stroke in new recommendations from the UK SunSafe program advo- Malmo, Sweden: the STROMA study. Neuroepidemiology 2005; 24:26–31.
cate that people should sun themselves often but for brief 10 Lejeune JP, Vinchon M, Amouyel P, Escartin T, Escartin D, intervals, without burning, preferably at mid-day, and without Christiaens JL. Association of occurrence of aneurysmal bleeding with sunscreen [27]. The US economic burden due to vitamin D meteorologic variations in the north of France. Stroke 1994; 25: 338– deficiency from inadequate exposure to solar UVB irradiation largely surpasses the cost of excess UV radiation [28]. This is 11 Brot C, Vestergaard P, Kolthoff N, Gram J, Hermann AP, Sorensen OH. Vitamin D status and its adequacy in healthy Danish largely due to the beneficial effects of vitamin D on the Ó 2009 International Society on Thrombosis and Haemostasis perimenopausal women: relationships to dietary intake, sun exposure patients with congestive heart failure: a double-blind, randomized, and serum parathyroid hormone. Br J Nutr 2001; 86(Suppl 1): S97– placebo-controlled trial. Am J Clin Nutr 2006; 83: 754–9.
21 Sedrani SH. Low 25-hydroxyvitamin D and normal serum calcium 12 Holick MF. Vitamin D deficiency. N Engl J Med 2007; 357: 266–81.
concentrations in Saudi Arabia: Riyadh region. Ann Nutr Metab 1984; 13 Lindqvist PG, Epstein E, Olsson H. The relationship between lifestyle factors and venous thromboembolism among women: a report from 22 Matsuoka LY, Ide L, Wortsman J, MacLaughlin JA, Holick MF.
the MISS study. Br J Haematol 2008; 144: 234–40.
Sunscreens suppress cutaneous vitamin D3 synthesis. J Clin Endocrinol 14 Tangpricha V, Turner A, Spina C, Decastro S, Chen TC, Holick MF.
Tanning is associated with optimal vitamin D status (serum 25-hy- 23 Clemens TL, Adams JS, Henderson SL, Holick MF. Increased skin droxyvitamin D concentration) and higher bone mineral density. Am pigment reduces the capacity of skin to synthesise vitamin D3. Lancet 15 Koyama T, Shibakura M, Ohsawa M, Kamiyama R, Hirosawa S.
24 Tsai AW, Cushman M, Rosamond WD, Heckbert SR, Polak JF, Anticoagulant effects of 1alpha,25-dihydroxyvitamin D3 on human Folsom AR. Cardiovascular risk factors and venous thromboembo- myelogenous leukemia cells and monocytes. Blood 1998; 92: 160–7.
lism incidence: the longitudinal investigation of thromboembolism 16 Sugden JA, Davies JI, Witham MD, Morris AD, Struthers AD.
etiology. Arch Intern Med 2002; 162: 1182–9.
Vitamin D improves endothelial function in patients with Type 2 25 Beer TM, Venner PM, Ryan CW, Petrylak DP, Chatta G, Dean diabetes mellitus and low vitamin D levels. Diabet Med 2008; 25: 320–5.
Ruether J, Chi KN, Curd JG, DeLoughery TG. High dose calcitriol 17 Scragg R, Sowers M, Bell C. Serum 25-hydroxyvitamin D, ethnicity, may reduce thrombosis in cancer patients. Br J Haematol 2006; 135: and blood pressure in the Third National Health and Nutrition Examination Survey. Am J Hypertens 2007; 20: 713–19.
26 Lindqvist P, Dahlba¨ck B, Marsa´l K. Thrombotic risk during preg- 18 Cannell JJ, Vieth R, Umhau JC, Holick MF, Grant WB, Madronich S, nancy: a population study. Obstet Gynecol 1999; 94: 595–9.
Garland CF, Giovannucci E. Epidemic influenza and vitamin D.
27 Gillie O. A new government policy is needed for sunlight and vita- Epidemiol Infect 2006; 134: 1129–40.
min D. Br J Dermatol 2006; 154: 1052–61.
19 Holick MF, Chen TC. Vitamin D deficiency: a worldwide problem 28 Grant WB, Garland CF, Holick MF. Comparisons of estimated with health consequences. Am J Clin Nutr 2008; 87: 1080S–6S.
economic burdens due to insufficient solar ultraviolet irradiance and 20 Schleithoff SS, Zittermann A, Tenderich G, Berthold HK, Stehle P, vitamin D and excess solar UV irradiance for the United States.
Koerfer R. Vitamin D supplementation improves cytokine profiles in Photochem Photobiol 2005; 81: 1276–86.
Ó 2009 International Society on Thrombosis and Haemostasis

Source: http://www.ultrasun.de/ims/editor/uploads/files/903%20Lindquist%20%20Sun%20habits%20and%20Venous%20Trombotic%20Embolism.pdf