Concept · medicine
Thyroid function and health outcomes
4 studiesEvidence last moved Sep 24, 2026
Thyroid function is usually read from two blood tests: thyroid-stimulating hormone (TSH) from the pituitary and free thyroxine (FT4) from the thyroid itself. Mild or 'subclinical' shifts in these values, and genetic findings in thyroid nodules, are increasingly linked to other health outcomes. This page covers a cohort of newborns exposed to dioxin through their mothers, cross-sectional and prospective studies linking thyroid values to dementia and macular degeneration in older adults, and a blinded cohort of RAS-mutated thyroid nodules.
Students often treat a lab value inside the 'normal range' as meaningless and a mutation as a diagnosis. The studies here show both assumptions can mislead. They also give practice in reading an association when design, sample size and timing limit what it can mean.
Studies
4
Findings
4
4 supporting · 0 challenging · 3 qualifying citations
Open tensions
1
Latest change
Concept page published
Thyroid function and health outcomes
Currently
What we know
- An environmental exposure can leave a measurable thyroid signal in the next generation.
- Where you sit inside the normal range can still matter for risk.
- A mildly overactive thyroid goes with dementia in older people, but the direction is unknown.
- A RAS mutation in a thyroid nodule is not the same as cancer, and the cancers it marks tend to be low-risk.
Largest unresolved question
Which thyroid measure carries the signal differs by outcome and design: in the Rotterdam cohort FT4 predicted macular degeneration while TSH did not, whereas in Sao Paulo the dementia association was defined by low TSH (with the lowest TSH quintile relevant only when below-range values were included). These are different outcomes, populations and designs, so they mark limits rather than a contradiction.
Common misconceptions
If TSH and FT4 are within the reference range, thyroid function is irrelevant to other disease.
In the Rotterdam cohort, higher-normal FT4 was associated with later macular degeneration, and the association held when restricted to people with fully normal thyroid function. It is an association, not proof of harm, but 'normal' is not the same as 'no gradient of risk'.
Finding a cancer-associated mutation such as RAS in a nodule means the nodule is cancer and needs aggressive treatment.
Fewer than half of RAS-positive nodules were malignant, the cancers were low-risk, and benign RAS-positive nodules followed for years did not grow significantly. Cytology caught every cancer, so RAS testing alone did not outperform the standard test.
An odds ratio of 4 for dementia shows subclinical hyperthyroidism causes dementia.
The Sao Paulo study measured thyroid status and dementia at the same time, so dementia or its causes could have altered thyroid function; with 33 exposed people the confidence interval ran from 1.3 to 13.1.
Related
Claim ledger
What the evidence shows
Drawn from 4 studies in this library. Mix labels say which citation roles are present; they are not a strength score. Supports means evidence for a finding; Challenges means evidence against a stated position; Qualifies marks scope.
An environmental exposure can leave a measurable thyroid signal in the next generation.
Maternal dioxin exposure from the 1976 Seveso accident was associated with raised newborn TSH decades later, with a gradient by contamination zone: 16.1% of babies of zone A mothers exceeded the 5-unit screening threshold versus 2.8% in the reference area (odds ratio 6.6).
- Can a mother's dioxin exposure decades earlier affect her baby's thyroid?— Raised screening TSH is a marker, not disease: only two babies had congenital hypothyroidism, and the dose-response with maternal plasma dioxin rested largely on five highly exposed mothers.
Where you sit inside the normal range can still matter for risk.
In a prospective Rotterdam cohort of 5,573 adults aged 55+, higher FT4 even within the normal range predicted incident age-related macular degeneration (highest vs middle quintile HR 1.34), while TSH did not.
- Is thyroid hormone level linked to macular degeneration?— Single baseline measurement, mostly white Europeans, and the overall quintile trend was not significant.
A mildly overactive thyroid goes with dementia in older people, but the direction is unknown.
Among 1,276 older adults in deprived areas of Sao Paulo, subclinical hyperthyroidism (low TSH, normal FT4) was associated with about four times the odds of dementia (OR 4.1, 95% CI 1.3-13.1), mainly vascular dementia; subclinical hypothyroidism showed no association.
- Is a mildly overactive thyroid linked to dementia in older adults?— Cross-sectional, only 33 exposed people and 49 dementia cases, adjusted only for age and BMI.
A RAS mutation in a thyroid nodule is not the same as cancer, and the cancers it marks tend to be low-risk.
In a blinded prospective cohort of 362 biopsied thyroid nodules, only 8 of 17 RAS-positive nodules (47%) were malignant, all were non-invasive encapsulated follicular-variant papillary cancers, and every cancer was also flagged by cytology.
Debates
Tensions and limits
Some items are genuine disagreements on the same question. Others mark different assays, populations, or outcomes.
Which thyroid measure carries the signal differs by outcome and design: in the Rotterdam cohort FT4 predicted macular degeneration while TSH did not, whereas in Sao Paulo the dementia association was defined by low TSH (with the lowest TSH quintile relevant only when below-range values were included). These are different outcomes, populations and designs, so they mark limits rather than a contradiction.
Which thyroid measure carries the signal differs by outcome and design: in the Rotterdam cohort FT4 predicted macular degeneration while TSH did not, whereas in Sao Paulo the dementia association was defined by low TSH (with the lowest TSH quintile relevant only when below-range values were included). These are different outcomes, populations and designs, so they mark limits rather than a contradiction.
- Is thyroid hormone level linked to macular degeneration?
- Is a mildly overactive thyroid linked to dementia in older adults?
Study Role Design N Population Outcome Is thyroid hormone level linked to macular degeneration? Supports CohortProspective population-based cohort (Rotterdam Study cohorts I and II); baseline TSH and free T4 related to incident AMD graded from repeated fundus photographs using Cox models, plus a bidirectional GWAS look-up for shared genetic variants. N=5573 · 5,573 participants aged 55 and over without AMD at baseline who had thyroid measurements and follow-up eye exams; 805 developed AMD. The genetic sub-analysis used 4,646 participants with genotype data. Mainly Caucasian adults aged 55 years and older living in Ommoord, a suburb of Rotterdam, the Netherlands. Incident early or late age-related macular degeneration (AMD) and specific AMD lesions (retinal pigment alterations, large drusen). Is a mildly overactive thyroid linked to dementia in older adults? Supports Cross-sectionalCross-sectional analysis of the baseline wave of a door-to-door population cohort, with logistic regression of dementia on thyroid status and TSH/FT4 quintiles, adjusted for age and BMI. N=1276 · 1,276 participants aged 65+ with thyroid tests and no overt thyroid disease or thyroid medication; of these, 33 had subclinical hyperthyroidism and 49 had dementia. Adults aged 65 and over living in deprived census areas (shantytowns and Family Health Program households) of Butantan, Sao Paulo, Brazil Presence of dementia (any type), Alzheimer's disease and vascular dementia diagnosed with the 10/66 Dementia Research Group protocol
PaperFren reads this as a limit on how far one study travels — different assays, populations, or outcomes — not a forced fight between papers.
Timeline
How understanding moved
Study years are when the paper was published. Evidence edits are dated changes to this page's claims. Explanations are when PaperFren added a Discovery — not a claim that the science happened that day.
2026
Concept page published
Thyroid function and health outcomes
Change log
What changed
Dated edits to this page's evidence: studies added or removed from a claim, claims added or withdrawn, and new explanations tagged here. Rewordings are not listed.
- Concept page published
Papers
4 studies in this library bear on Thyroid function and health outcomes, ordered by citations.
- Can a mother's dioxin exposure decades earlier affect her baby's thyroid?
Babies born up to three decades after the Seveso dioxin accident to mothers from the most contaminated area had clearly higher thyroid-stimulating hormone at birth, rising step by step with contamination level.
- Is thyroid hormone level linked to macular degeneration?
Older adults with higher levels of free thyroid hormone, even within the normal range, were more likely to develop age-related macular degeneration over the following years.
- Does a RAS gene mutation mean a thyroid lump is cancer?
Thyroid nodules carrying a RAS mutation turned out to be cancer in fewer than half of cases, and the cancers that were found were all low-risk.
- Is a mildly overactive thyroid linked to dementia in older adults?
Older Brazilians whose blood tests showed a mildly overactive thyroid were about four times as likely to have dementia, mainly vascular dementia, as those with normal thyroid function.
Compare studies
Select 2–10 studies. Design and N are labels, not a ranking.
Nothing selected yet.
Questions
What is still open
Which thyroid measure carries the signal differs by outcome and design: in the Rotterdam cohort FT4 predicted macular degeneration while TSH did not, whereas in Sao Paulo the dementia association was defined by low TSH (with the lowest TSH quintile relevant only when below-range values were included). These are different outcomes, populations and designs, so they mark limits rather than a contradiction.
Ask PaperFren about Thyroid function and health outcomes
Study this conceptflashcards and short-answer questions
Compare the strength of evidence linking thyroid function to macular degeneration and to dementia in the studies on this page.
The macular degeneration evidence comes from a prospective population cohort of 5,573 older Dutch adults in which baseline FT4 preceded the outcome, with 805 incident cases over about 7 years; the highest normal FT4 quintile had HR 1.34. The dementia evidence comes from a cross-sectional survey of 1,276 older Brazilians with only 33 people with subclinical hyperthyroidism, so temporal order is unknown and the interval is wide (OR 4.1, 1.3-13.1). The cohort design gives better protection against reverse causation, though both remain observational and neither measured thyroid function more than once.
What does the Seveso newborn TSH study show, and what does it not show?
It shows that babies born 1994-2005 to women living in the most contaminated zone at the 1976 dioxin accident had higher screening TSH, with a gradient across zones (16.1% vs 2.8% above threshold) and a positive correlation with maternal plasma dioxin. It does not show clinical harm, since only two babies had congenital hypothyroidism and development was not followed. The dose-response correlation depended on five very highly exposed mothers, and maternal iodine intake was not measured.