Sunburn is the exposure to avoid, and it is nearly the whole of the well-evidenced advice. UV clearly raises the risk of melanoma, and the pattern that does the damage is intermittent, intense, burning exposure on unprotected skin. Everything else about the sun is a genuine trade-off, and the evidence behind its two sides is very uneven. A randomized trial and a gold meta-analysis anchor the skin-cancer harm. The much-repeated claim that sun-seekers live longer rests on a single observational cohort of Swedish women, wide open to reverse causation. Oral vitamin D does not stand in for whatever the sun might do, because supplements do not move all-cause mortality. So the verdict is narrow and stratum-dependent: avoid burning, and beyond that, judge how much sun is worth it by your skin type and latitude.
Skin cancer: the pattern is the exposure, not the dose
For melanoma, the pattern of UV exposure matters more than the total dose, and reading risk off total dose hides the finding. Gandini’s meta-analysis (57 studies, 38,671 cases) splits the exposure into patterns that point in opposite directions (Gandini et al., 2005):
- A history of sunburn carries the largest risk: RR 2.03 (95% CI 1.73-2.37).
- Intermittent, recreational exposure raises risk: RR 1.61 (1.31-1.99).
- Total exposure of all kinds is a weak signal: RR 1.34 (1.02-1.77).
- Chronic, occupational exposure is null, even slightly inverse: RR 0.95 (0.87-1.04).
The split is not an artifact of comparing different study sets. On the 19 studies that reported both patterns, intermittent exposure (RR 1.46) exceeded chronic (RR 1.09), P=0.015 (Gandini et al., 2005). Sunburn risk climbs with latitude (RR 2.54 at 50 degrees or higher) and with fair skin, and childhood sunburn carries RR 2.24. A funnel-plot asymmetry suggests the sunburn estimate is inflated, but even the trim-and-fill correction (RR 1.73) leaves a clear positive (Gandini et al., 2005).
Chronic exposure’s null result can mislead, and Gandini flags it directly: it does not mean occupational sun protects. The comparison group for chronic exposure is low-continuous exposure, which itself sweeps in both high-intermittent people and people who get almost no sun at all — a confounded contrast, not a protective effect (Gandini et al., 2005).
The mechanism explains why pattern beats dose. An intermittent burst of UV hits pale, unaccustomed skin and reaches the melanocytes; regular exposure falls on tanned, thickened skin that blocks more of it at the epidermis (Gandini et al., 2005). That is why the sharpest marker is sunburn — intense UV actually reaching the cells that turn malignant. The estimates are observational, so recall bias and control selection apply, and the analysis treated moles as an intermediate rather than adjusting them away.
Melanoma is the lethal endpoint the numbers above track, but it is not the only skin cancer UV drives. Nonmelanoma skin cancers (basal- and squamous-cell) also rise with sun exposure (Lindqvist et al., 2016); they are far more common but rarely fatal and are usually treated locally, and the fabric does not hold their effect sizes — a named gap.
Sunscreen is the one randomized protection, and it worked without cutting sun time
The Nambour trial is the only randomized test of sunscreen against melanoma. Green randomized 1,621 adults in a predominantly fair-skinned Australian town (Nambour) to daily SPF-16 versus discretionary use from 1992 to 1996 and followed them for about 15 years (Green et al., 2011). All melanoma fell to HR 0.50 (0.24-1.02, P=.051) — a borderline result, 11 cases in 812 people (1.3%) against 22 in 809 (2.7%). Invasive melanoma fell further, HR 0.27 (0.08-0.97), a 73% reduction, though the authors label the invasive subgroup exploratory. In-situ melanoma did not move (HR 0.73, 0.29-1.81).
Two features matter for how to read it. Outdoor time was similar in both arms, so the benefit did not come from the sunscreen group simply avoiding the sun — it came from blocking UV while still outside. And it is a single trial with small event counts and a lead author who disclosed L’Oreal research funding, so it is suggestive, not settled.
This changes the target. The big lever for melanoma is avoiding intermittent burning, not minimizing total time in the sun, and sunburn history is both the largest risk and the most modifiable one. Early-life exposure matters too — migrant studies show risk tracks childhood sun, so burns in childhood carry weight that adult caution cannot fully undo. The risk concentrates in fair skin and at high latitude or high ambient UV, which is where the same behaviour costs the most.
Gandini and Green are complementary, not independent confirmation of one number. Gandini establishes that an intermittent UV pattern is associated with melanoma; Green shows that blocking UV reduces it. But an observational risk ratio and a randomized intervention effect are different quantities — you cannot pool them. Their agreement is a convergence of two designs on one direction, not two independent estimates of a single effect. Together they make UV’s causal, modifiable role in melanoma hard to dismiss.
The eye is a second, smaller UV harm
Skin cancer is not the only patient-important harm from UV. UV is an established risk factor for age-related cataract, the leading global cause of blindness, alongside diabetes, higher BMI, hypertension, smoking, and steroid use. The one meta-analysis held here (Aune) names UV among those risk factors but measured physical activity against cataract, not UV, so it flags the harm without sizing it (Aune et al., 2026). No held source quantifies a UV-to-cataract dose, so this arm cannot be weighed against the melanoma numbers — but it points the same way and licenses one near-free move: eye protection under strong sun (Cataract and Modifiable Exposures).
All-cause mortality: one Swedish cohort, wide open to reverse causation
The counterweight to the skin-cancer harm is the claim that people who seek the sun live longer. It comes almost entirely from one study. Lindqvist followed 29,518 Swedish women for 20 years and found that those who avoided the sun had roughly 1.7 to 2 times the all-cause mortality of sun-seekers; the gap widened dose-dependently for cardiovascular death, and avoiders lost an estimated 0.6 to 2.1 years of life expectancy (Lindqvist et al., 2016). Taken at face value, that is a large signal. It should not be taken at face value.
The people who avoid the sun in this cohort are not a random slice of it. They are older, poorer, less educated, far more likely to be sick at baseline (comorbidity 20.4% versus 6.7%), and less physically active than the sun-seekers. Sick and frail people avoid the sun — which is the textbook setup for reverse causation and confounding by frailty, where poor health lowers the exposure rather than the exposure protecting health (The U-Shaped Association Artifact). The association did survive adjustment for age, smoking, education, income, comorbidity, and exercise, which raises its credibility without settling it. The authors are explicit about the limit (Lindqvist et al., 2016):
«it is not possible to differentiate between active sun exposure habits and a healthy lifestyle, and secondly, the results are of an observational nature; therefore, a causal link cannot be proven.»
No randomized trial has tested a sun-seeking versus sun-avoiding strategy against mortality, and none is likely to. The widely-quoted line that sun avoidance is a mortality risk on the scale of smoking only compares life-expectancy estimates within this one cohort; it does not demonstrate a causal equivalence. Under symmetric standards, a counterintuitive observational signal earns no pass for being surprising, exactly as the conventional melanoma harm earns none for being expected. The mortality benefit is plausible and unreplicated; the melanoma harm is randomized and pooled. Do not treat the two as equally established.
Vitamin D: the mechanism everyone reaches for, and why a pill does not reproduce it
The intuitive explanation for a sun-mortality benefit is vitamin D: sun makes it, low levels track worse health, so more sun should help. But the pill does not reproduce the benefit in people who are already replete. VITAL randomized 25,871 US adults to vitamin D3 2000 IU per day and found no effect on cancer incidence (HR 0.96, 0.88-1.06) or major cardiovascular events (HR 0.97, 0.85-1.12) (Manson et al., 2019); its ancillary fracture trial added no effect on total fractures (HR 0.98, 0.89-1.08) (LeBoff et al., 2022). Zhang pooled far more evidence — 50 randomized trials (74,655 participants, GRADE high): vitamin D did not lower all-cause mortality, RR 0.98 (0.95-1.02) (Zhang et al., 2019).
One endpoint is not null, and symmetric standards require naming it: the pill does seem to lower cancer death. VITAL put cancer mortality at HR 0.83 (0.67-1.02) (Manson et al., 2019), and Zhang’s pooled estimate reached significance for cancer death (RR 0.85, 0.74-0.97, I2=0%) (Zhang et al., 2019). That signal is real, but it does not rescue the sun-mortality story. The sun claim is about all-cause mortality, and that stays flat; a supplement that trims cancer death without moving the all-cause total is not the mechanism behind a claimed survival benefit.
This is the surrogate discipline in action (Surrogate Outcomes). A vitamin-D blood level is a marker; it earns the status of a treatment target only if raising it by supplement actually moves an outcome, and on all-cause mortality it does not. If the sun’s mortality signal ran through vitamin D, then supplementing vitamin D should reproduce it — and it fails to. Lindqvist reaches the same conclusion from the other direction (Lindqvist et al., 2016):
«Whether the positive effect of sun exposure demonstrated in this observational study is mediated by vitamin D, another mechanism related to UV radiation, or by unmeasured bias cannot be determined from our results. Vitamin D levels might be just a marker of sun exposure.»
The supplement trials enrolled largely replete populations, so they test whether adding vitamin D to someone who already has enough buys anything (it does not) — a different question from correcting a genuine deficiency (Deficiency Repletion vs Enhancement). That distinction matters for the deficient, but it does not rescue the sun-mortality story. It means the pill is not the mechanism. If the sun does anything for mortality, it more likely runs through non-vitamin-D UV pathways — nitric oxide lowering blood pressure, melatonin, skin beta-endorphin — none of which a supplement supplies. So oral vitamin D does not substitute for sun on this question.
The trade-off: the same habit sits on both arms
The tempting resolution is to split the exposure — say that intermittent burning causes melanoma while calm, regular sun lowers mortality, so one strategy captures the benefit and dodges the harm. The sources do not support that escape. Lindqvist built the sun-exposure score entirely from recreational, intermittent items: summer sunbathing, winter and mountain sunbathing, holidays abroad to sunbathe, and tanning-bed use (Lindqvist et al., 2016). That is precisely Gandini’s harmful intermittent pattern, and tanning beds are themselves an established melanoma risk. So the same sun-seeking behaviour raises melanoma in Gandini and associates with lower mortality in Lindqvist. This is a real trade-off on one exposure, not two exposures that conveniently separate.
Three consequences follow for a person deciding what to do:
- The one exposure that might sit well on both arms — regular, genuinely non-burning sun — is a hypothesis, not a measured sweet spot. Lindqvist’s score cannot tell non-burning habitual sun apart from recreational burning, so do not present it as an evidenced safe zone.
- Avoiding sunburn is the robust, well-evidenced move, but on this evidence it does not come free of the mortality-arm exposure, because the two are tangled in the same recreational habit.
- Sunscreen is the partial escape: in Nambour it cut melanoma without cutting outdoor time, so it lets a person keep sun exposure while lowering the burning that carries the risk. It is not, on Lindqvist’s caution, a licence to overexpose.
Skin type and latitude move the balance, mostly through baseline melanoma risk
The net verdict shifts by stratum, and most of the shift is baseline-risk arithmetic (Baseline Risk and the Relative-Absolute Split). The relative melanoma effect is roughly constant, but the same relative effect is a much larger absolute harm where melanoma is common:
- Fair skin (Fitzpatrick I-II) in high ambient UV — northern Australia, a year-round UV index of 6 or more — puts melanoma harm on top. Protect, and do not burn.
- Darker skin, or high latitude with low UV — Sweden, where the UV index sits below 3 for eight or nine months — carries a far lower melanoma baseline. The same relative harm is then a much smaller absolute one, and the weak, confounded mortality signal carries relatively more weight. This is where Lindqvist argues that blanket year-round sun avoidance may not help.
The weighting past that point is the person’s to make. Melanoma risk, the uncertain mortality signal, and the quality-of-life value of time outdoors sit on one health axis, and no formula collapses them into a single number. A person at high melanoma risk should protect; a person at low melanoma risk in a low-UV country has more room, and the evidence does not force either of them into a corner.
Note one off-axis effect and stop: sun exposure drives skin photoageing and cosmetic damage. That is real, but it is not a health outcome, it carries no data here, and it should not be priced against the melanoma or mortality findings. Name it, and let the person weigh it themselves.
What to do
- Do not burn. This is the big lever and it is nearly free: sunburn carries the largest melanoma signal (RR ~2), and childhood sunburn is worst of all, so protecting children and skipping holiday burning bouts buys most of the achievable risk reduction.
- Use sunscreen and cover up in strong sun. It is the one protective step a randomized trial showed lowers melanoma, and it worked without anyone spending less time outdoors.
- Protect the eyes too. Sunglasses or a brimmed hat in strong sun cost almost nothing, and UV is an established cataract risk factor — a distinct harm on the vision side.
- Do not treat a vitamin-D pill as a stand-in for sunlight. Supplement trials in replete people show no all-cause mortality benefit. Correcting a genuine deficiency is a separate question (Deficiency Repletion vs Enhancement).
- Weigh skin type and latitude. Fair skin under strong year-round sun: prioritize protection. Darker skin at high latitude: strict year-round avoidance is not clearly warranted, and the modest sun most people get is not the melanoma driver — burning is.
- Do not chase the sun makes you live longer headline. It rests on one unreplicated, confounded cohort, and it attaches to the same recreational habit that raises melanoma.
What the evidence does not settle
The loop here is open. No source measures the net effect of a sun-seeking versus sun-avoiding strategy against all-cause mortality in a trial, so the central trade-off is adjudicated by judgment, not by a clean experiment. The specific gaps worth naming:
- The mortality benefit rests on one cohort of Swedish women. A second independent mortality cohort, or a Mendelian-randomization analysis that fixes lifetime UV or vitamin-D exposure, would test whether the Lindqvist signal replicates outside that narrow stratum. Neither is held.
- The candidate sweet spot — regular, non-burning sun — has never been measured as its own exposure. Lindqvist’s recreational score cannot isolate it, so it stays a hypothesis.
- Circadian and light-exposure effects on sleep and mood are a plausible non-UV channel for daytime outdoor light, but the wiki holds no source on them here, so they are a named gap, not a finding.
- Sauna and heat exposure are sometimes folded into sunlight advice; they are a separate exposure with their own single-cohort evidence base and belong to a different question (Sauna Bathing and Cardiovascular Mortality), not to the UV trade-off.
Avoiding sunburn is the move the strongest evidence supports. The claimed mortality benefit of sun-seeking is observational, confounded, and unreplicated — plausible but far from established — and a vitamin-D pill does not stand in for it. Where the balance lands beyond do not burn depends on skin type, latitude, and what the person values, and the evidence supports saying exactly that rather than manufacturing a number.
Evidence box
Question ’What does the evidence show about sun / UV exposure across its patient-important outcomes — skin cancer (melanoma and keratinocyte cancers) against the all-cause-mortality and vitamin-D signals — how does each effect vary with dose, skin type, and latitude, and how should a person weigh the trade-off?‘ Evidence included 7 sources — 3 gold, 4 high Overall certainty Low-moderate (see Rating Certainty of Evidence) Source-selection note All sources are gold or high tier. Last updated 2026-10-01 · Independently reviewed: No · Full edit history