NetCortex

Is that health headline true? Check it in five moves

Written by AI from saved sources. A script checks every quote and every figure against the saved sources, and a second AI compares each claim with its source. How we work and corrections

WHO means World Health Organization. Its cancer agency, IARC, put aspartame in Group 2B: possibly cancer-causing. The intake limit stayed unchanged.

This page shows how to check a headline like this yourself, in five moves.

Aspartame and cancer.

  1. Read past the headline.
  2. Find what the expert groups concluded.
  3. Look for a review.
  4. Check who paid, and when.
  5. Match the answer to the headline.

Checked 27 October 2026.

Why trust the experts at all?

Experts can be paid. So why trust them? Yes, doubt can be paid for. A publicly funded 2002 investigation uncovered tobacco companies’ hidden efforts to challenge secondhand-smoke research.

A systematic review examines all relevant research to answer one question. It uses a set method and stated rules for what to include. The National Academies of Sciences, Engineering, and Medicine publish reports written by expert committees. Their “Consensus Study Reports” record an expert committee’s consensus based on evidence. IARC’s full group reviews its subgroups’ drafts and reaches a shared evaluation.

IARC checks declared interests before inviting participants and may limit their role. Invited specialists with conflicts cannot write sections describing or interpreting cancer data. They cannot take part in the evaluations. WHO says JECFA’s aspartame panel was screened and found free of conflicts of interest.

Commercial sources with a real or possible stake in the findings cannot fund or conduct Cochrane reviews. Authors must declare their interests when they submit a review or its written plan. Employees of companies with a real or possible financial interest in the result cannot be authors.

Financial gain may influence professional judgment about research. Cochrane’s 2017 review compared drug and device studies funded by industry with those funded from other sources. Industry-funded studies more often reached conclusions favouring their products, though the evidence was low quality. Low quality, now called low certainty, means limited confidence in the estimated effect. The true effect may be substantially different.

Cochrane says combining and comparing studies makes its reviews more reliable than individual studies. So a group weighing all the evidence, with declared interests, is harder to sway than one expert or one study. The fourth move shows how to check who paid. In 2015 experiments, two groups of 40 university students rated repeated statements as truer than new ones, even when they knew better.

Five moves, using that headline

Before move 1: find the report behind the headline

Find the report or study named in the article. Note who wrote it and the year. For CNN’s headline, find the July 2023 IARC and JECFA aspartame assessments. IARC’s expert group and JECFA’s panel made the assessments. Open their joint summary. Without a link, search IARC JECFA aspartame assessment 2023.

Decide: If the article names no report, study or expert body, treat the headline as unchecked and do not share it yet.

  1. Read past the headline

    Do this: Read the CNN article’s opening paragraphs. Ask: Can aspartame cause cancer? Then ask: How likely is harm at the amount consumed?

    Our result: The article reports “possibly carcinogenic”, meaning possibly cancer-causing. It says the label does not mean your diet soda causes cancer. The International Agency for Research on Cancer (IARC) assessed the first question. The Joint Expert Committee on Food Additives (JECFA) assessed the second.

    Decide: Read beyond “possible cancer cause” before sharing. If the article discusses amounts consumed, take both questions into the next move.

  2. Find what the expert groups concluded

    Scientific consensus means broad agreement among experts in the relevant field. It is based on how they read the available evidence. It does not require 100% agreement and has no fixed percentage.

    For aspartame, IARC and JECFA reviewed the evidence on potential cancer hazard and other health risks. Do this: Type aspartame cancer site:who.int into Google. The site: part limits results to that website. Open “Aspartame hazard and risk assessment results released”. Here, hazard means potential to harm. Risk means how likely harm is. Find “JECFA concluded”. If Google shows nothing, search the WHO news page by keyword and year.

    Our result: In an announcement published on 14 July 2023, IARC reported classifying aspartame as Group 2B: possibly cancer-causing. This was based on limited evidence. For cancer in humans, limited evidence means a link was seen. But other explanations could not be confidently ruled out. This describes how strong the cancer evidence is, not your chance of cancer from a particular amount. In that announcement, WHO reported JECFA’s finding: there was “no sufficient reason to change” its intake limit. JECFA reaffirmed that consumption within its daily limit is safe. Its acceptable daily intake is an estimated amount of an additive in food or drinking water. That amount can be safely consumed daily over a lifetime without harmful effects.

    For another view, type aspartame site:fda.gov into Google; open Aspartame and Other Sweeteners in Food. The US Food and Drug Administration (FDA) disagrees that the studies support IARC’s label. It reports no safety concerns under approved conditions of use. If Google shows nothing, use the search in the FDA newsroom. For another topic, try the national health agency’s own site search.

    For other topics, try the US National Academies search. Replace aspartame with your topic. If the search shows nothing, type aspartame site:nationalacademies.org into Google. On a report’s page, find the “Consensus Study Report” label. Read its conclusions.

    Decide: If bodies answer different questions, keep their answers separate. If they disagree on the same question, report the disagreement. For our headline, retain both the possible cancer label and the unchanged intake limit.

  3. Look for a review of the same question

    Do this: Open this PubMed search, or type aspartame cancer into its search box. PubMed searches medical and life-science research. Choose “Systematic Review” in the article-type filters; use “See all article type filters” if needed. Read a relevant result’s summary and conclusion.

    Cochrane is a group that brings health studies together to answer research questions. Alternatively, open Cochrane Library, choose “Search”, type aspartame, and select “Cochrane Reviews”. If nothing fits, try PubMed’s “Systematic Review” filter.

    Our result: A 2025 review in Advances in Nutrition covered 90 observational studies of sweeteners including aspartame. Researchers watch what people are exposed to and what diseases they get, rather than influencing either. It found no consistent link with cancer overall. But people reported their own intake, and other possible causes were not always fully taken into account. Keep it for the funding check below.

    A 2020 Cochrane sweetener review concerns people with diabetes. Trials compare the effects of one treatment with another. The trials tested sweeteners for four to ten months. That does not answer the lifetime cancer question.

    Decide: Keep the review’s conclusion and limits together. Use it only if it answers the headline’s question. If nothing fits, return to move 2. Try another search before concluding that no evidence exists.

  4. Check who paid, and when

    Do this: Open the 2025 cancer review. Use find-in-page for “Funding”, then “Conflict of interest”. A conflict of interest means another interest, such as money, may affect judgment about research. Read who paid and any account of their involvement. In the Cochrane review, find “Published”, then “up-to-date”. Here, up-to-date tells you the date of the evidence search included in the review.

    Our result: The review of sweeteners, including aspartame, names the American Beverage Association as its funder. The association represents America’s non-alcoholic drink industry. The Cochrane review was published on 25 May 2020. But it used evidence up to May 2019.

    Decide: If the funder benefits from the conclusion, seek a separate assessment before relying on that review alone. A financial link does not automatically mean a paper was improperly influenced. If a government agency paid for and wrote the report, seek an outside review too. If the evidence is older than the headline, seek a later assessment before calling it the latest answer. Compare reviews from several independent groups. Check who paid. Do not rely on familiarity.

  5. Match the answer to the headline

    Do this: Return to the WHO announcement. Find “limited evidence”, then “no sufficient reason to change”. Compare both with the headline. Ask what each number counts and what amount was studied. Compare that with the amount people actually get.

    For a second example, the National Toxicology Program (NTP) reviewed total fluoride exposure in 2024. It covered all sources, not drinking water alone. With moderate confidence, it linked higher exposure, such as water above 1.5 milligrams per litre, to lower IQ in children. The US recommended water level is 0.7 milligrams per litre. There were insufficient data to tell whether that level harms children’s IQ. In 2020, the National Academies reviewed an unfinished version of the NTP fluoride report. They checked the report’s use of evidence, without conducting their own independent evaluation of the evidence.

    Decide: If a study examines higher exposure than people usually get, check for evidence at lower exposure before applying its findings to everyday use.

    Our answer: The headline correctly reports a possible cancer label and an unchanged intake limit. It does not establish that ordinary consumption causes cancer. WHO said safety was not a major concern at commonly used doses. It also called for more and better studies.

    Decide: If you share this story, include both halves of the headline and the uncertainty. If a post turns possibly into proven, do not share it as established fact. Use this wording: The 2023 assessment gave aspartame a possible cancer label on limited evidence. The intake limit stayed unchanged.

What about 97% agreement?

Check it yourself

Suppose a post says: 97% of all scientists agree about human-caused warming. Does the 2013 study support that wording?

Show the answer

Ask: 97% of what? In a 2013 climate study, 97.1% of paper summaries taking a position supported human-caused warming. Most summaries took no position; this was not a poll of every scientist. Keep that qualification when sharing the number.

Lookalikes to check

  • A petition: ask whether signers research this subject. The Oregon Petition’s minimum requirement for signers was a bachelor’s degree in science. In 1998, the National Academy of Sciences said the petition had nothing to do with it.
  • A famous scientist: a degree does not ensure expert knowledge in another highly technical field.
  • A workshop record: National Academies workshop records report what participants said. Opinions in these records do not carry the institution’s backing.
  • One voice on each side: giving equal space to a vocal critic with an extreme stance and a subject expert representing scientific consensus can mislead. It can create a false impression of scientific division.

Can consensus change?

Yes: stress and lifestyle were once considered the main causes of peptic ulcers. Research established a link with bacteria called H. pylori. Moving continents also faced resistance before new ocean-floor evidence helped establish the case. Look for the new evidence behind a challenge.

Before you share

  1. Have I read beyond the headline? Yes / No
  2. Have I found a relevant expert report or review? Yes / No
  3. Have I checked the date and who paid? Yes / No
  4. Do I know what any number counts? Yes / No
  5. Does my version keep the source’s limits and disagreements? Yes / No

If any answer is no, pause sharing and return to that move.

Worked examples on this site

Sources

CNN, aspartame headline (updated 14 July 2023); WHO, joint announcement (14 July 2023) and food additives; IARC, classification questions and answers; FDA, aspartame response; Cochrane, review methods, industry sponsorship review (2017) and diabetes review (2020); sweeteners and cancer review (2025); tobacco documents investigation (2002); repetition experiments (2015); journal disclosure guidance; The Journalist’s Resource; National Academies publication types; climate-paper study (2013); Oregon Petition qualifications; academy petition statement (1998); scientific expertise; Nobel Assembly on ulcers; continental-drift history; PubMed search instructions; Cochrane search tutorial; Google search help; NTP fluoride report (2024); WHO news search; FDA’s full name; FDA newsroom; National Academies review of the NTP report (2020); NHS clinical-trial definition; Cochrane evidence-certainty definitions; Cochrane search-date guidance; Cochrane search-date policy (2019); National Academies review request (2020); NHS ulcer definition; IARC, expert selection and assessment methods; Cochrane, funding and author-interest rules; WHO, joint aspartame summary (2023); IARC, aspartame assessment group (2023).