"Sunscreen is more dangerous due to chemical absorption and vitamin D blocking than moderate sun exposure."
Five independent authoritative sources — from dermatology, oncology, and peer-reviewed research — all reject this claim, and no credible evidence supports it.
What Was Claimed?
The claim is that wearing chemical sunscreen is actually riskier than skipping it and getting some sun. The argument goes that sunscreen ingredients seep into your bloodstream and could be toxic, and that blocking UV rays prevents your body from making vitamin D — and that these two dangers combined make sunscreen worse for you than moderate unprotected sun exposure.
This kind of claim circulates widely on social media and wellness circles, often cited by people trying to make informed choices about their health. It raises legitimate questions worth examining.
What Did We Find?
The absorption part of the claim is grounded in real science — but stops well short of supporting the conclusion. FDA-sponsored studies published in JAMA (2019 and 2020) did confirm that chemical UV filters, including oxybenzone, are absorbed into the bloodstream at concentrations exceeding FDA thresholds. Oxybenzone was detected at up to 258 ng/mL and remained in the body for up to 21 days. Those are real findings. But the studies' own authors explicitly concluded that this "does not indicate that individuals should refrain from the use of sunscreen." Absorption is documented. Harm from that absorption has not been demonstrated.
The endocrine disruption concern — that oxybenzone acts like a hormone disruptor — comes from animal studies, but those studies used exposure levels equivalent to roughly 277 years of daily human sunscreen application. Human volunteer studies at realistic exposure levels showed no biologically significant changes in reproductive hormones. Oxybenzone has been in use since 1978 with no demonstrated human carcinogenicity.
On vitamin D: laboratory studies do show that high-SPF sunscreen reduces UVB-driven vitamin D synthesis under controlled conditions. But a 2022 international expert panel reviewed real-world population data — including people on outdoor vacations — and found no significant difference in vitamin D levels between sunscreen users and non-users with equivalent time spent outdoors. Incomplete application, residual UV exposure, and atmospheric factors all compensate. And vitamin D can be reliably obtained from food and supplements, making "skip the sunscreen for vitamin D" an unnecessary trade-off.
The most direct rebuttal is simply the scale of what UV radiation actually does. About 90% of nonmelanoma skin cancers are caused by UV exposure from the sun. High-quality randomized controlled trial evidence shows that sunscreen use reduces melanoma risk by approximately 50% and squamous cell carcinoma by approximately 40%. The claim's premise — that sunscreen is more dangerous than UV — is directly inverted by the evidence.
When researchers looked at data suggesting sunscreen users had higher cancer rates, the explanation turned out to be behavioral: people who use more sunscreen tend to spend more time in the sun. The sunscreen use was a marker of sun-seeking behavior, not a cause of cancer.
What Should You Keep In Mind?
The uncertainty around chemical absorption is real and ongoing — the FDA has called for more research, and that research hasn't finished. "Not proven harmful yet" is not the same as "proven safe." If this concerns you, mineral sunscreens (zinc oxide, titanium dioxide) are not absorbed into the bloodstream and are a well-supported alternative.
The vitamin D question is genuinely nuanced in controlled settings. People with very limited outdoor exposure and poor dietary intake could in theory be affected, though population studies don't bear this out in practice.
What the evidence does not support is the comparative claim — that skipping sunscreen and accepting UV exposure is the safer choice. The harms of UV radiation are well-documented and substantial; the claimed harms of sunscreen have not been demonstrated at realistic human exposure levels.
How Was This Verified?
This claim was evaluated by searching for authoritative sources on each stated mechanism, conducting four adversarial searches specifically looking for evidence that could support the claim, and verifying all citations against live web pages. Full details are in the structured proof report and the full verification audit. To inspect the logic and reproduce the result, see re-run the proof yourself.
What could challenge this verdict?
Four independent adversarial searches were conducted before writing this proof:
1. FDA absorption studies (potential support for the claim): FDA-sponsored JAMA studies from 2019 and 2020 confirm that oxybenzone and other chemical UV filters are absorbed into the bloodstream at concentrations exceeding FDA thresholds, with oxybenzone detected up to 258 ng/mL and persisting 21 days. This is the strongest factual basis for the claim. However, both studies' authors explicitly stated "these findings do not indicate that individuals should refrain from the use of sunscreen," and the FDA's response was to call for more research — not to restrict sunscreen use. This adversarial finding does not break the disproof; it identifies the strongest factual premise and shows it stops short of supporting the harm conclusion.
2. Oxybenzone endocrine disruption (potential support for the claim): Animal studies do suggest oxybenzone could act as an endocrine disruptor. However, these studies used exposure concentrations equivalent to ~277 years of daily human sunscreen application. Human volunteer studies showed no biologically significant alterations in reproductive hormones at realistic exposure levels. Oxybenzone has been in use since 1978 with no demonstrated human carcinogenicity or endocrine harm. Does not break the proof.
3. Vitamin D deficiency in real-world sunscreen users: No evidence was found of clinically significant vitamin D deficiency caused by sunscreen use in real-world populations. The 2022 PMC expert panel found no difference in vitamin D levels between sunscreen users and non-users with equivalent outdoor exposure. Does not break the proof.
4. Peer-reviewed studies supporting the claim directly: No peer-reviewed study was found concluding that sunscreen use causes greater net harm than moderate unprotected sun exposure. The apparent correlation in some observational data (sunscreen users with higher cancer rates) is explained by behavioral confounding: people who use more sunscreen spend more time in the sun. The causal direction runs from sun exposure to cancer, with sunscreen use as an imperfect marker of sun-seeking behavior. Does not break the proof.
Source: author analysis
Sources
| Source | ID | Type | Verified |
|---|---|---|---|
| American Academy of Dermatology — Sunscreen FAQs | B1 | Unclassified | Yes |
| MD Anderson Cancer Center — Sunscreen Myths Debunked | B2 | Unclassified | Yes |
| PMC / CMAJ — Efficacy and Safety of Sunscreen (2020) | B3 | Government | Yes |
| PMC — Sunscreen and Vitamin D Expert Panel (2022) | B4 | Government | Yes |
| Skin Cancer Foundation — Skin Cancer Facts & Statistics | B5 | Unclassified | Yes |
| Verified disproof source count | A1 | — | Computed |
detailed evidence
Evidence Summary
| ID | Fact | Verified |
|---|---|---|
| B1 | American Academy of Dermatology: absorbed ingredients not proven harmful | Yes |
| B2 | MD Anderson Cancer Center: no medical evidence sunscreen causes cancer; UV does | Yes |
| B3 | PMC/CMAJ 2020: high-quality evidence sunscreen reduces melanoma and nonmelanoma cancer | Yes |
| B4 | PMC 2022 expert panel: sunscreen does not limit vitamin D production in real-world use | Yes |
| B5 | Skin Cancer Foundation: ~90% of nonmelanoma skin cancers associated with UV exposure | Yes |
| A1 | Verified disproof source count | Computed: 5 verified disproof sources out of 5 consulted (threshold: 3) |
Source: proof.py JSON summary
Proof Logic
The claim rests on two stated causal mechanisms and a comparative harm conclusion. Each is addressed in turn.
Mechanism 1 — Chemical absorption causes harm (B1, B2): FDA-sponsored research does confirm that chemical sunscreen ingredients including oxybenzone, avobenzone, octocrylene, homosalate, octisalate, and octinoxate are absorbed into the bloodstream at concentrations exceeding the FDA's threshold of concern (0.5 ng/mL). Oxybenzone was detected at up to 258 ng/mL and remained measurable for 21 days after application. This establishes that absorption occurs. However, absorption alone does not establish harm. The FDA called for additional studies but did not conclude any ingredient is unsafe. The AAD is explicit: "Just because an ingredient is absorbed into the bloodstream does not mean that it is harmful or unsafe" (B1). MD Anderson concurs: "There is no medical evidence that sunscreen causes cancer" (B2). Studies suggesting oxybenzone could act as an endocrine disruptor used exposure concentrations equivalent to approximately 277 years of daily sunscreen application — far beyond any realistic human exposure.
Mechanism 2 — Vitamin D blocking causes harm (B4): Laboratory-controlled studies do show that high-SPF sunscreen can reduce UVB-driven vitamin D synthesis. However, the 2022 PMC international expert panel reviewed real-world population data and found that sunscreen users and non-users with equivalent outdoor exposure showed no significant difference in vitamin D levels (B4). Incomplete real-world application, residual UVB exposure, and ozone-layer attenuation of the shortest UVB wavelengths all compensate. Vitamin D can also be reliably obtained from dietary sources and supplements, making avoidance of sunscreen an unnecessary and counterproductive strategy for this purpose.
Comparative harm conclusion — sunscreen vs. UV (B3, B5): Even setting aside the absence of demonstrated harm from sunscreen, the comparative framing collapses against the scale of documented UV-caused harm. About 90% of nonmelanoma skin cancers (basal cell and squamous cell carcinoma) are attributable to UV radiation (B5). Approximately 86% of melanomas are similarly attributed. High-quality randomized controlled trial evidence — including a major trial showing 40% fewer squamous cell carcinomas in the sunscreen arm — establishes that sunscreen is protective, not harmful (B3). The claim's comparative harm direction is inverted relative to the evidence.
Source independence (A1): All five sources are from distinct institutions: a professional medical association (AAD), a major cancer research and treatment hospital (MD Anderson), peer-reviewed academic literature published in CMAJ (PMC/CMAJ 2020), an international expert consensus panel (PMC 2022), and an independent cancer prevention nonprofit (Skin Cancer Foundation). No shared authorship or institutional affiliation. Verified source count: 5 ≥ 3 (threshold).
Source: author analysis
Conclusion
Verdict: DISPROVED
5 of 5 independently sourced, institutionally diverse, authoritative sources reject the claim, exceeding the threshold of 3. All citations were fully verified via live page fetch.
The claim's two stated mechanisms fail on their own terms: (1) chemical absorption is documented but has not been shown to cause harm at realistic human exposure levels; (2) vitamin D reduction from sunscreen is a laboratory finding that does not translate to meaningful deficiency in real-world outdoor populations. On the comparative harm dimension, the claim is directly inverted: UV radiation causes 86–90% of all skin cancers, and sunscreen use is backed by randomized trial evidence reducing melanoma risk by ~50% and squamous cell carcinoma by ~40%.
No adversarial evidence was found that breaks the disproof. The claim is not supported by peer-reviewed evidence and is explicitly rejected by the American Academy of Dermatology, MD Anderson Cancer Center, Skin Cancer Foundation, and the peer-reviewed academic literature indexed on PubMed Central.
Note: B1 (aad.org), B2 (mdanderson.org), and B5 (skincancer.org) are classified as tier 2 (unclassified) by the automated domain credibility system. These institutions are among the most authoritative in their respective domains; the tier-2 classification reflects only that their domains are not in the automated allowlist, not any concern about their credibility. The two PMC citations (B3, B4) are independently sufficient to support the disproof at the threshold of 3 — the conclusion does not rest solely on tier-2 sources.
audit trail
All 5 citations verified.
Original audit log
B1 — American Academy of Dermatology - Status: verified - Method: full_quote - Fetch mode: live - Coverage: N/A (full match)
B2 — MD Anderson Cancer Center - Status: verified - Method: full_quote - Fetch mode: live - Coverage: N/A (full match)
B3 — PMC / CMAJ (2020) - Status: verified - Method: full_quote - Fetch mode: live - Coverage: N/A (full match)
B4 — PMC Sunscreen & Vitamin D Expert Panel (2022) - Status: verified - Method: full_quote - Fetch mode: live - Coverage: N/A (full match)
B5 — Skin Cancer Foundation - Status: verified - Method: full_quote - Fetch mode: live - Coverage: N/A (full match)
All five citations verified via full quote match on live page fetch. No citation failures. No impact notes required.
Source: proof.py JSON summary
| Field | Value |
|---|---|
| Subject | Chemical sunscreen |
| Property | Comparative harm vs. moderate sun exposure |
| Operator | >= |
| Proof direction | disprove |
| Threshold | 3 verified authoritative sources rejecting the claim |
| Operator note | The claim asserts sunscreen is MORE dangerous than moderate sun exposure, citing two mechanisms: (1) systemic absorption of chemical UV filters and (2) reduction of vitamin D synthesis. Interpreted as: the combined harm of sunscreen use >= the harm of moderate unprotected sun exposure. We DISPROVE this by gathering authoritative sources showing: (a) chemical absorption is documented but no human harm has been demonstrated; (b) real-world vitamin D reduction from sunscreen is minimal — population studies show no significant difference in vitamin D levels between sunscreen users and non-users with equivalent outdoor exposure; (c) UV radiation is the dominant cause of skin cancer (~86-90% of cases); and (d) sunscreen use reduces melanoma risk by ~50% and squamous cell carcinoma by ~40%. Threshold: 3 or more independently verified authoritative sources must reject the claim's conclusion for a DISPROVED verdict. |
Source: proof.py JSON summary
Natural-language claim: "Sunscreen is more dangerous due to chemical absorption and vitamin D blocking than moderate sun exposure."
Formal interpretation: The claim asserts that the combined harm of sunscreen use (from two stated mechanisms) is greater than or equal to the harm from moderate unprotected sun exposure. This requires that: 1. Chemical absorption of UV filter ingredients causes meaningful human harm, AND 2. Vitamin D synthesis reduction from sunscreen use causes meaningful human harm, AND 3. Together, these harms exceed the well-documented harms of UV exposure (skin cancer, premature aging).
Operator choice: The comparison "more dangerous than" is interpreted as harm(sunscreen) > harm(moderate sun). We disprove the claim by showing that the stated harms of sunscreen are either non-existent or negligible at real-world exposure levels, while the harms of UV radiation are well-documented and substantial. Threshold for DISPROVED verdict: 3 or more independently verified authoritative sources must reject the claim's conclusion. A threshold of 3 was chosen because this is a broad medical safety claim where consensus across diverse authoritative institutions is required. Using a lower threshold of 2 would be insufficient for a claim with significant public health implications.
Source: proof.py JSON summary
| Fact ID | Domain | Type | Tier | Note |
|---|---|---|---|---|
| B1 | aad.org | unknown | 2 | Unclassified domain — verify source authority manually. The American Academy of Dermatology is the largest dermatological professional organization in the United States and a primary standard-setter for dermatological care. |
| B2 | mdanderson.org | unknown | 2 | Unclassified domain — verify source authority manually. MD Anderson Cancer Center is a National Cancer Institute-designated Comprehensive Cancer Center and one of the leading cancer research and treatment institutions in the world. |
| B3 | nih.gov | government | 5 | Government domain (.gov). PubMed Central (PMC) is the open-access archive of the U.S. National Institutes of Health. The underlying article is published in CMAJ (Canadian Medical Association Journal), a peer-reviewed medical journal. |
| B4 | nih.gov | government | 5 | Government domain (.gov). PubMed Central (PMC); underlying article is an international expert consensus panel review. |
| B5 | skincancer.org | unknown | 2 | Unclassified domain — verify source authority manually. The Skin Cancer Foundation is the leading international organization dedicated to skin cancer prevention, with an independent photobiology advisory committee. |
Source: proof.py JSON summary
Note on tier-2 sources: The automated credibility system classifies domains by allowlist membership, not institutional authority. The three tier-2 sources (AAD, MD Anderson, Skin Cancer Foundation) are among the most authoritative institutions in dermatology and oncology. Their tier-2 classification reflects only that their domains are not in the automated allowlist. The disproof conclusion is independently supported by the two tier-5 (NIH/PMC) sources (B3, B4), which alone provide two of the three required verified sources.
Source: author analysis
Date: 2026-03-31 (System date matches proof generation date)
Verifying citations...
[✓] source_aad: Full quote verified for source_aad (source: tier 2/unknown)
[✓] source_mdanderson: Full quote verified for source_mdanderson (source: tier 2/unknown)
[✓] source_pmc_cmaj: Full quote verified for source_pmc_cmaj (source: tier 5/government)
[✓] source_pmc_vitd: Full quote verified for source_pmc_vitd (source: tier 5/government)
[✓] source_skincancer: Full quote verified for source_skincancer (source: tier 2/unknown)
Confirmed sources: 5 / 5
verified disproof sources vs threshold: 5 >= 3 = True
Source: proof.py inline output (execution trace)
| Description | Sources Consulted | Sources Verified |
|---|---|---|
| Multiple independent authoritative institutions consulted | 5 | 5 |
Verification status by source:
| Source Key | Status |
|---|---|
| source_aad | verified |
| source_mdanderson | verified |
| source_pmc_cmaj | verified |
| source_pmc_vitd | verified |
| source_skincancer | verified |
Independence note: Sources span distinct institution types: professional medical association (AAD), major cancer research and treatment center (MD Anderson), peer-reviewed academic literature (PMC/CMAJ 2020), international expert consensus panel (PMC 2022), and an independent cancer prevention foundation (Skin Cancer Foundation). All five are institutionally independent with no shared authorship. The two PMC citations (B3, B4) alone exceed the threshold of 3 when combined with any one other verified source.
Source: proof.py JSON summary
Check 1 — FDA absorption studies as potential support for the claim - Question: Does the FDA's own research confirm that sunscreen chemicals are absorbed, potentially supporting the danger claim? - Verification performed: Searched for 'FDA sunscreen chemical absorption oxybenzone bloodstream JAMA 2019 2020'. Found two FDA-sponsored JAMA studies (2019 and 2020) confirming that oxybenzone and other ingredients are absorbed into the bloodstream at concentrations exceeding the FDA threshold of 0.5 ng/mL. Oxybenzone reached up to 258.1 ng/mL and remained detectable for up to 21 days. However, both study authors explicitly stated: 'These findings do not indicate that individuals should refrain from the use of sunscreen.' The FDA called for more research but did not conclude harm. - Finding: FDA studies confirm absorption is real (oxybenzone up to 258 ng/mL, persisting 21 days), but the studies' own authors explicitly state this does not mean sunscreen should be avoided. Absorption documented; harm from that absorption: not demonstrated. - Breaks proof: No
Check 2 — Oxybenzone endocrine disruption in humans - Question: Is there evidence that oxybenzone or other absorbed chemicals disrupt hormones or cause cancer in humans at realistic exposure levels? - Verification performed: Searched for 'oxybenzone endocrine disruptor humans evidence cancer realistic exposure'. Harvard Health and NYP/Weill Cornell both report that studies showing hormone disruption used concentrations equivalent to approximately 277 years of daily sunscreen application. Human volunteer studies confirmed no biologically significant alterations in reproductive hormones at real-world exposure levels. Oxybenzone has been in use since 1978 with no demonstrated human carcinogenicity. - Finding: Animal-model endocrine effects require ~277 years of equivalent daily human use. Human studies show no biologically significant reproductive hormone changes. No human carcinogenicity has been demonstrated after decades of use. - Breaks proof: No
Check 3 — Clinically significant vitamin D deficiency from sunscreen - Question: Does sunscreen use lead to clinically significant vitamin D deficiency in real-world populations? - Verification performed: Searched for 'sunscreen vitamin D deficiency real world population studies clinical'. PMC 2022 expert panel found that population studies of outdoor individuals (vacationers) showed vitamin D levels 'did not seem to differ between those applying sunscreen and those who did not, with exposure time and body surface area exposed being equivalent.' Real-world sunscreen application is typically incomplete; shorter UVB wavelengths that drive vitamin D synthesis are also attenuated by the ozone layer, partially compensating. The Skin Cancer Foundation recommends dietary sources and supplements as the safe route to vitamin D rather than UV exposure. - Finding: Real-world population studies show no significant difference in vitamin D levels between sunscreen users and non-users given equivalent outdoor exposure time. Laboratory-controlled reductions do not translate to clinically meaningful deficiency in practice. Dietary supplementation is the recommended and safe vitamin D source. - Breaks proof: No
Check 4 — Peer-reviewed support for the direct claim - Question: Are there peer-reviewed studies directly concluding that sunscreen is more dangerous than moderate unprotected sun exposure on a net-harm basis? - Verification performed: Searched for 'sunscreen more dangerous sun exposure net harm peer reviewed evidence'. Harvard Health explains that studies where sunscreen users appeared to have higher cancer rates reflect behavioral confounding: people who use more sunscreen tend to spend more time in the sun, reversing the causal arrow. MD Anderson, AAD, and the Skin Cancer Foundation all explicitly reject the conclusion that sunscreen causes net harm exceeding sun exposure. No peer-reviewed study was found concluding sunscreen use causes greater net harm than moderate unprotected sun exposure. - Finding: No peer-reviewed studies support this conclusion. The observed correlation is reversed causation: higher-sunscreen users get more sun, not cancer from sunscreen. All major medical institutions — AAD, MD Anderson, Skin Cancer Foundation, CDC — explicitly reject the claim. - Breaks proof: No
Source: proof.py JSON summary
| Rule | Status | Notes |
|---|---|---|
| Rule 1: Every empirical value parsed from quote text, not hand-typed | N/A — qualitative proof; no numeric values extracted from quotes | Qualitative consensus proof; citation status is the "value" |
| Rule 2: Every citation URL fetched and quote checked | PASS | All 5 citations verified via verify_all_citations() — full_quote match, live fetch |
| Rule 3: System time used for date-dependent logic | PASS | date.today() used; PROOF_GENERATION_DATE cross-checked against system date |
| Rule 4: Claim interpretation explicit with operator rationale | PASS | CLAIM_FORMAL with operator_note documents both mechanisms, comparative harm interpretation, and threshold rationale |
| Rule 5: Adversarial checks searched for independent counter-evidence | PASS | 4 adversarial searches conducted: FDA absorption studies, oxybenzone endocrine disruption, vitamin D deficiency in populations, direct peer-reviewed support |
| Rule 6: Cross-checks used independently sourced inputs | PASS | 5 sources from distinct institutions (AAD, MD Anderson, CMAJ, PMC expert panel, Skin Cancer Foundation) — no shared authorship |
| Rule 7: Constants and formulas imported from computations.py, not hand-coded | PASS | compare() used for claim evaluation; no hard-coded constants or inline formulas |
| validate_proof.py | PASS | 15/15 checks passed, 0 issues, 0 warnings |
Source: author analysis
For this qualitative/consensus disproof proof, extraction records capture citation verification status rather than numeric values.
| Fact ID | Extracted Value (Citation Status) | Counts Toward Threshold | Quote Snippet (first 80 chars) |
|---|---|---|---|
| B1 | verified | Yes | "Just because an ingredient is absorbed into the bloodstream does not mean that i" |
| B2 | verified | Yes | "There is no medical evidence that sunscreen causes cancer. However, there is a l" |
| B3 | verified | Yes | "High-quality evidence has shown that sunscreen reduces the risk of developing bo" |
| B4 | verified | Yes | "Sunscreens can be effective in preventing erythema from solar exposure without l" |
| B5 | verified | Yes | "About 90 percent of nonmelanoma skin cancers are associated with exposure to ult" |
Extraction method: Citation verification status (verified/partial/not_found/fetch_failed) is assigned by verify_all_citations() in scripts/verify_citations.py. For qualitative proofs, this status serves as the "extracted value" — a source counts toward the disproof threshold if and only if its quote was found on the live page (status = "verified" or "partial"). All five sources returned "verified" via full_quote match on live fetch.
Source: proof.py JSON summary + author analysis
Cite this proof
Proof Engine. (2026). Claim Verification: “Sunscreen is more dangerous due to chemical absorption and vitamin D blocking than moderate sun exposure.” — Disproved. https://doi.org/10.5281/zenodo.19489799
Proof Engine. "Claim Verification: “Sunscreen is more dangerous due to chemical absorption and vitamin D blocking than moderate sun exposure.” — Disproved." 2026. https://doi.org/10.5281/zenodo.19489799.
@misc{proofengine_sunscreen_is_more_dangerous_due_to_chemical_absorption_and_vitamin_d_blocking,
title = {Claim Verification: “Sunscreen is more dangerous due to chemical absorption and vitamin D blocking than moderate sun exposure.” — Disproved},
author = {{Proof Engine}},
year = {2026},
url = {https://proofengine.info/proofs/sunscreen-is-more-dangerous-due-to-chemical-absorption-and-vitamin-d-blocking/},
note = {Verdict: DISPROVED. Generated by proof-engine v1.3.1},
doi = {10.5281/zenodo.19489799},
}
TY - DATA TI - Claim Verification: “Sunscreen is more dangerous due to chemical absorption and vitamin D blocking than moderate sun exposure.” — Disproved AU - Proof Engine PY - 2026 UR - https://proofengine.info/proofs/sunscreen-is-more-dangerous-due-to-chemical-absorption-and-vitamin-d-blocking/ N1 - Verdict: DISPROVED. Generated by proof-engine v1.3.1 DO - 10.5281/zenodo.19489799 ER -
View proof source
This is the exact proof.py that was deposited to Zenodo and runs when you re-execute via Binder. Every fact in the verdict above traces to code below.
"""
Proof: Sunscreen is more dangerous due to chemical absorption and vitamin D blocking
than moderate sun exposure.
Generated: 2026-03-31
"""
import json
import os
import sys
PROOF_ENGINE_ROOT = os.environ.get("PROOF_ENGINE_ROOT")
if not PROOF_ENGINE_ROOT:
_d = os.path.dirname(os.path.abspath(__file__))
while _d != os.path.dirname(_d):
if os.path.isdir(os.path.join(_d, "proof-engine", "skills", "proof-engine", "scripts")):
PROOF_ENGINE_ROOT = os.path.join(_d, "proof-engine", "skills", "proof-engine")
break
_d = os.path.dirname(_d)
if not PROOF_ENGINE_ROOT:
raise RuntimeError("PROOF_ENGINE_ROOT not set and skill dir not found via walk-up from proof.py")
sys.path.insert(0, PROOF_ENGINE_ROOT)
from datetime import date
from scripts.verify_citations import verify_all_citations, build_citation_detail
from scripts.computations import compare
# ── Rule 3: Anchor to system time ──────────────────────────────────────────────
PROOF_GENERATION_DATE = date(2026, 3, 31)
actual = date.today()
if actual == PROOF_GENERATION_DATE:
today = PROOF_GENERATION_DATE
date_note = "System date matches proof generation date"
else:
today = actual
date_note = f"Proof generated for {PROOF_GENERATION_DATE}, running on {actual}"
print(f" Date: {today} ({date_note})")
# ── Rule 4: Explicit claim interpretation ──────────────────────────────────────
CLAIM_NATURAL = (
"Sunscreen is more dangerous due to chemical absorption and vitamin D blocking "
"than moderate sun exposure."
)
CLAIM_FORMAL = {
"subject": "Chemical sunscreen",
"property": "comparative harm vs. moderate sun exposure",
"operator": ">=",
"operator_note": (
"The claim asserts sunscreen is MORE dangerous than moderate sun exposure, "
"citing two mechanisms: (1) systemic absorption of chemical UV filters and "
"(2) reduction of vitamin D synthesis. Interpreted as: the combined harm of "
"sunscreen use >= the harm of moderate unprotected sun exposure. "
"We DISPROVE this by gathering authoritative sources showing: "
"(a) chemical absorption is documented but no human harm has been demonstrated; "
"(b) real-world vitamin D reduction from sunscreen is minimal — population "
"studies show no significant difference in vitamin D levels between sunscreen "
"users and non-users with equivalent outdoor exposure; "
"(c) UV radiation is the dominant cause of skin cancer (~86-90% of cases); and "
"(d) sunscreen use reduces melanoma risk by ~50% and squamous cell carcinoma "
"by ~40%. Threshold: 3 or more independently verified authoritative sources "
"must reject the claim's conclusion for a DISPROVED verdict."
),
"threshold": 3,
"proof_direction": "disprove",
}
# ── Fact Registry ──────────────────────────────────────────────────────────────
FACT_REGISTRY = {
"B1": {
"key": "source_aad",
"label": "American Academy of Dermatology: absorbed ingredients not proven harmful",
},
"B2": {
"key": "source_mdanderson",
"label": "MD Anderson Cancer Center: no medical evidence sunscreen causes cancer; UV does",
},
"B3": {
"key": "source_pmc_cmaj",
"label": "PMC/CMAJ 2020: high-quality evidence sunscreen reduces melanoma and nonmelanoma cancer",
},
"B4": {
"key": "source_pmc_vitd",
"label": "PMC 2022 expert panel: sunscreen does not limit vitamin D production in real-world use",
},
"B5": {
"key": "source_skincancer",
"label": "Skin Cancer Foundation: ~90% of nonmelanoma skin cancers associated with UV exposure",
},
"A1": {
"label": "Verified disproof source count",
"method": None,
"result": None,
},
}
# ── Empirical Facts — sources that REJECT the claim (confirm it is false) ─────
# (Rule 5: adversarial sources supporting the claim go in adversarial_checks only)
empirical_facts = {
"source_aad": {
"quote": (
"Just because an ingredient is absorbed into the bloodstream does not mean "
"that it is harmful or unsafe."
),
"url": "https://www.aad.org/media/stats-sunscreen",
"source_name": "American Academy of Dermatology — Sunscreen FAQs",
},
"source_mdanderson": {
"quote": (
"There is no medical evidence that sunscreen causes cancer. However, there is "
"a lot of evidence that UV rays from the sun and tanning beds do."
),
"url": "https://www.mdanderson.org/cancerwise/sunscreen-myths-debunked.h00-159697545.html",
"source_name": "MD Anderson Cancer Center — Sunscreen Myths Debunked",
},
"source_pmc_cmaj": {
"quote": (
"High-quality evidence has shown that sunscreen reduces the risk of developing "
"both melanoma and nonmelanoma skin cancer."
),
"url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC7759112/",
"source_name": "PMC / CMAJ — Efficacy and Safety of Sunscreen (2020)",
},
"source_pmc_vitd": {
"quote": (
"Sunscreens can be effective in preventing erythema from solar exposure without "
"limiting the benefits with respect to vitamin D production."
),
"url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC9002342/",
"source_name": "PMC — Sunscreen and Vitamin D Expert Panel (2022)",
},
"source_skincancer": {
"quote": (
"About 90 percent of nonmelanoma skin cancers are associated with exposure to "
"ultraviolet (UV) radiation from the sun."
),
"url": "https://www.skincancer.org/skin-cancer-information/skin-cancer-facts/",
"source_name": "Skin Cancer Foundation — Skin Cancer Facts & Statistics",
},
}
# ── Rule 2: Citation verification ─────────────────────────────────────────────
print("\nVerifying citations...")
citation_results = verify_all_citations(empirical_facts, wayback_fallback=True)
# ── Count sources with verified citations ─────────────────────────────────────
COUNTABLE_STATUSES = ("verified", "partial")
n_confirmed = sum(
1 for key in empirical_facts
if citation_results[key]["status"] in COUNTABLE_STATUSES
)
print(f" Confirmed sources: {n_confirmed} / {len(empirical_facts)}")
# ── Rule 7: Use compare() — never hardcode claim_holds ────────────────────────
claim_holds = compare(
n_confirmed,
CLAIM_FORMAL["operator"],
CLAIM_FORMAL["threshold"],
label="verified disproof sources vs threshold",
)
# ── Rule 5: Adversarial checks — search for evidence SUPPORTING the claim ─────
adversarial_checks = [
{
"question": (
"Does the FDA's own research confirm that sunscreen chemicals are absorbed, "
"potentially supporting the danger claim?"
),
"verification_performed": (
"Searched for 'FDA sunscreen chemical absorption oxybenzone bloodstream JAMA 2019 2020'. "
"Found two FDA-sponsored JAMA studies (2019 and 2020) confirming that oxybenzone and "
"other ingredients are absorbed into the bloodstream at concentrations exceeding the "
"FDA threshold of 0.5 ng/mL. Oxybenzone reached up to 258.1 ng/mL and remained "
"detectable for up to 21 days. However, both study authors explicitly stated: "
"'These findings do not indicate that individuals should refrain from the use of "
"sunscreen.' The FDA called for more research but did not conclude harm."
),
"finding": (
"FDA studies confirm absorption is real (oxybenzone up to 258 ng/mL, persisting "
"21 days), but the studies' own authors explicitly state this does not mean sunscreen "
"should be avoided. Absorption documented; harm from that absorption: not demonstrated."
),
"breaks_proof": False,
},
{
"question": (
"Is there evidence that oxybenzone or other absorbed chemicals disrupt hormones "
"or cause cancer in humans at realistic exposure levels?"
),
"verification_performed": (
"Searched for 'oxybenzone endocrine disruptor humans evidence cancer realistic exposure'. "
"Harvard Health and NYP/Weill Cornell both report that studies showing hormone disruption "
"used concentrations equivalent to approximately 277 years of daily sunscreen application. "
"Human volunteer studies confirmed no biologically significant alterations in reproductive "
"hormones at real-world exposure levels. Oxybenzone has been in use since 1978 with no "
"demonstrated human carcinogenicity."
),
"finding": (
"Animal-model endocrine effects require ~277 years of equivalent daily human use. "
"Human studies show no biologically significant reproductive hormone changes. "
"No human carcinogenicity has been demonstrated after decades of use."
),
"breaks_proof": False,
},
{
"question": (
"Does sunscreen use lead to clinically significant vitamin D deficiency "
"in real-world populations?"
),
"verification_performed": (
"Searched for 'sunscreen vitamin D deficiency real world population studies clinical'. "
"PMC 2022 expert panel found that population studies of outdoor individuals (vacationers) "
"showed vitamin D levels 'did not seem to differ between those applying sunscreen and "
"those who did not, with exposure time and body surface area exposed being equivalent.' "
"Real-world sunscreen application is typically incomplete; shorter UVB wavelengths that "
"drive vitamin D synthesis are also attenuated by the ozone layer, partially compensating. "
"The Skin Cancer Foundation recommends dietary sources and supplements as the safe route "
"to vitamin D rather than UV exposure."
),
"finding": (
"Real-world population studies show no significant difference in vitamin D levels between "
"sunscreen users and non-users given equivalent outdoor exposure time. Laboratory-controlled "
"reductions do not translate to clinically meaningful deficiency in practice. "
"Dietary supplementation is the recommended and safe vitamin D source."
),
"breaks_proof": False,
},
{
"question": (
"Are there peer-reviewed studies directly concluding that sunscreen is more dangerous "
"than moderate unprotected sun exposure on a net-harm basis?"
),
"verification_performed": (
"Searched for 'sunscreen more dangerous sun exposure net harm peer reviewed evidence'. "
"Harvard Health explains that studies where sunscreen users appeared to have higher cancer "
"rates reflect behavioral confounding: people who use more sunscreen tend to spend more time "
"in the sun, reversing the causal arrow. MD Anderson, AAD, and the Skin Cancer Foundation "
"all explicitly reject the conclusion that sunscreen causes net harm exceeding sun exposure. "
"No peer-reviewed study was found concluding sunscreen use causes greater net harm than "
"moderate unprotected sun exposure."
),
"finding": (
"No peer-reviewed studies support this conclusion. The observed correlation is reversed "
"causation: higher-sunscreen users get more sun, not cancer from sunscreen. All major "
"medical institutions — AAD, MD Anderson, Skin Cancer Foundation, CDC — explicitly "
"reject the claim."
),
"breaks_proof": False,
},
]
# ── Verdict and structured output ─────────────────────────────────────────────
if __name__ == "__main__":
any_unverified = any(
cr["status"] != "verified" for cr in citation_results.values()
)
is_disproof = CLAIM_FORMAL.get("proof_direction") == "disprove"
any_breaks = any(ac.get("breaks_proof") for ac in adversarial_checks)
if any_breaks:
verdict = "UNDETERMINED"
elif claim_holds and not any_unverified:
verdict = "DISPROVED" if is_disproof else "PROVED"
elif claim_holds and any_unverified:
verdict = (
"DISPROVED (with unverified citations)"
if is_disproof
else "PROVED (with unverified citations)"
)
elif not claim_holds:
verdict = "UNDETERMINED"
else:
verdict = "UNDETERMINED"
FACT_REGISTRY["A1"]["method"] = (
f"count(verified disproof citations) = {n_confirmed}"
)
FACT_REGISTRY["A1"]["result"] = str(n_confirmed)
citation_detail = build_citation_detail(FACT_REGISTRY, citation_results, empirical_facts)
# For qualitative proofs: extractions record citation verification status
extractions = {}
for fid, info in FACT_REGISTRY.items():
if not fid.startswith("B"):
continue
ef_key = info["key"]
cr = citation_results.get(ef_key, {})
extractions[fid] = {
"value": cr.get("status", "unknown"),
"value_in_quote": cr.get("status") in COUNTABLE_STATUSES,
"quote_snippet": empirical_facts[ef_key]["quote"][:80],
}
summary = {
"fact_registry": {
fid: {k: v for k, v in info.items()}
for fid, info in FACT_REGISTRY.items()
},
"claim_formal": CLAIM_FORMAL,
"claim_natural": CLAIM_NATURAL,
"citations": citation_detail,
"extractions": extractions,
"cross_checks": [
{
"description": "Multiple independent authoritative institutions consulted",
"n_sources_consulted": len(empirical_facts),
"n_sources_verified": n_confirmed,
"sources": {k: citation_results[k]["status"] for k in empirical_facts},
"independence_note": (
"Sources span distinct institution types: professional medical association "
"(AAD), major cancer research and treatment center (MD Anderson), "
"peer-reviewed academic literature (PMC/CMAJ 2020), international expert "
"consensus panel (PMC 2022), and an independent cancer prevention foundation "
"(Skin Cancer Foundation). All five are institutionally independent with no "
"shared authorship."
),
}
],
"adversarial_checks": adversarial_checks,
"verdict": verdict,
"key_results": {
"n_confirmed": n_confirmed,
"threshold": CLAIM_FORMAL["threshold"],
"operator": CLAIM_FORMAL["operator"],
"claim_holds": claim_holds,
},
"generator": {
"name": "proof-engine",
"version": open(os.path.join(PROOF_ENGINE_ROOT, "VERSION")).read().strip(),
"repo": "https://github.com/yaniv-golan/proof-engine",
"generated_at": date.today().isoformat(),
},
}
print("\n=== PROOF SUMMARY (JSON) ===")
print(json.dumps(summary, indent=2, default=str))
Re-execute this proof
The verdict above is cached from when this proof was minted. To re-run the exact
proof.py shown in "View proof source" and see the verdict recomputed live,
launch it in your browser — no install required.
Re-execute the exact bytes deposited at Zenodo.
Re-execute in Binder runs in your browser · ~60s · no installFirst run takes longer while Binder builds the container image; subsequent runs are cached.
machine-readable formats
Downloads & raw data
found this useful? ★ star on github