Iodine is the textbook deficiency nutrient — its lower arm (deficiency -> impaired child neurodevelopment) is held on Deficiency Repletion vs Enhancement. This page holds the upper arm: chronic excess iodine intake raises thyroid-disease risk, principally hypothyroidism, in iodine-replete free-living populations. Different outcome, different stratum, different decision — so iodine is the nutrient for which the wiki now characterises both arms of one dose-response U, each on its own outcome. The single source is a gold-design SR+MA whose underlying evidence is low quality — a distinction that governs how far the finding carries.
The decision this changes: in a population already iodine-sufficient, more is not better — monitor drinking-water and salt iodine, do not over-supplement, and watch the vulnerable (autoimmune / antibody-positive / nodular / thyroidectomised) strata. (Katagiri et al., 2017) — the excess->hypothyroidism finding and the gold-design/low-underlying-quality split are Katagiri’s; — the whole-U framing (this page as the upper arm to Bougma’s lower arm) and the “more is not better” decision gloss are the wiki’s.
The effect estimate
- Population / comparator: free-living adults in areas of chronic iodine excess (median urinary iodine concentration, UIC, above the WHO cutoff) vs iodine-adequate areas. Meta-analysis was pooled over cross-sectional studies only (3 adult studies for each pooled outcome). (Katagiri et al., 2017)
- Outcome (patient-important, with a surrogate caveat): overt hypothyroidism (OH) and subclinical hypothyroidism (SCH). SCH is the load-bearing signal; it is a biochemical state (raised TSH, normal free T4) part-way to a patient-important outcome, so treat it as a target-not-outcome -> Surrogate Outcomes. Katagiri’s own warrant for reporting it: some SCH is treated and some is preventable. (Katagiri et al., 2017)
- Magnitude (relative): adults, excess vs adequate — OH OR 2.78 (95% CI 1.47 to 5.27); SCH OR 2.03 (1.58 to 2.62). Dropping the one elderly study: OH 2.44 (1.21 to 4.91), SCH 1.95 (1.47 to 2.58) — the estimate is not carried by the nursing-home cohort. (Katagiri et al., 2017)
- Absolute: not poolable here — the constituent studies mostly reported crude, unadjusted percentages, and baseline SCH prevalence varied widely by area, so no single baseline risk anchors an absolute effect. This is a stated limit, not an omission. (inferred from Katagiri et al., 2017)
- Certainty (the source’s own): «Because of the low quality and limited number of included studies, further evidence and review are required.» (Katagiri et al., 2017)
Dose-response — direction is solid, the threshold is edge-of-evidence
- Where the cutoff comes from. «the WHO’s epidemiologic criteria define a median UIC 300 μg/L as ‘Excessive’ in a population of school-age children», applied to adults (500 μg/L for pregnant / lactating); the US Tolerable Upper Intake Level is 1100 μg/day. (Katagiri et al., 2017) The 300 μg/L number is a definitional WHO cutpoint, not a located knee in the harm curve.
- The studied range is far above the cutoff — so the threshold is unestimated. In the cross-sectional adult studies the excess areas ran UIC 338 to 1241 μg/L: «Since the high iodine areas in many studies showed a UIC which was much higher than the cut-off value of iodine excess, we were unable to clarify the situation in borderline excess areas.» (Katagiri et al., 2017) Below ~340 μg/L there is essentially no data; an apparent optimum at 300 is the sampling edge, not a curve feature.
- Gradient present. Across the plotted studies «these studies show an increase in the percentage of SCH along with an increase in median UIC» (Katagiri et al., 2017); one large child water-iodine study reported a goitre gradient rising with dose (RR 1.56 at 300-399, 2.27 at 600-699, 3.46 at 900-999, 3.69 above 1500 μg/L, referent 100-200). (Katagiri et al., 2017)
- An interventional anchor. A double-blind RCT (Sang) «proposed that a total intake of 800 μg/ day confers a risk of SCH (supplementation 400 μg, median UIC 672 μg)» in people whose baseline intake was already excess. (Katagiri et al., 2017)
Mechanism — Wolff-Chaikoff escape failure
«In most individuals, escape from the Wolff-Chaikoff effect caused by an acute excess occurs due to a decrease in sodium-iodide symporter (NIS) activity … In some individuals in whom high residual NIS activity prevents adaptation to the Wolff-Chaikoff effect, iodine excess induces hypothyroidism.» (Katagiri et al., 2017) The mechanism is human-corroborated and gives the excess->hypothyroidism direction a physiological warrant, not merely an observational association — though the «mechanism … is not completely obvious.» (Katagiri et al., 2017)
The other tail is iodine-induced hyperthyroidism (Jod-Basedow / IIH): a supplementation side effect concentrated where fortification was recently introduced into previously deficient areas, or in nodular individuals. In chronic-excess areas this review found only a non-significant hyperthyroid increase — so the durable chronic-excess signal is hypothyroid, not hyperthyroid. (inferred from Katagiri et al., 2017)
Who is vulnerable — the effect-modification stratum (route b/c)
Predisposing thyroid damage is the susceptibility axis: «individuals with predisposing thyroid damage such as autoimmune thyroiditis or thyroidectomy are susceptible to iodine-induced hypothyroidism.» (Katagiri et al., 2017) Antibody status modifies the effect: «excess iodine intake and positive TPO antibody were risk factors for SCH independently … and interaction effects on the incidence of SCH between high UIC and positive antibody were found in children.» (Katagiri et al., 2017)
This is a genuine effect-modification / contraindication stratum (routes b/c on the transportability table): the harm fires hardest in antibody-positive / autoimmune-thyroiditis / nodular / post-thyroidectomy individuals. But the review excluded antibody studies by design, so the interaction is signalled, not quantified — a named gap it flags for a dedicated review. (inferred from Katagiri et al., 2017)
Is the upper arm real, or a U-shaped artifact?
Run the The U-Shaped Association Artifact discipline. This is not a protective-lower-arm case — it is an upper arm of harm — so the question is whether the harm arm survives a check before it is believed, and it does, directionally:
- Interventional corroboration. Sang’s double-blind RCT raises TSH / induces SCH dose-dependently when the already-replete are supplemented — a randomised design is the strongest form of the interventional check the decision rule asks for, and it is not vulnerable to the reverse-causation / confounding that bends observational arms.
- Mechanism + gradient. Wolff-Chaikoff escape failure (human-corroborated) plus the UIC-SCH gradient make the direction coherent, not a bare shape.
So believe the direction (chronic excess raises hypothyroidism risk, chiefly SCH, chiefly in the vulnerable). Discount the magnitude: the pooled ORs come from cross-sectional, crude, unadjusted percentages with ecological (area-level) exposure contrasts and unremoved antibody/age/sex confounding — «many studies only identified the crude percentages of the diseases and did not adjust these variables.» (Katagiri et al., 2017) No referent-correction and no Mendelian-randomization on the arm, so the size stays low-certainty even though the sign is corroborated. (inferred from Katagiri et al., 2017)
Measurement caveat — UIC is a population marker, not an intake
The exposure is urinary iodine concentration, a population-level status marker, not measured individual intake — spot UIC is noisy at the individual level, area medians can be pulled by intermittent high-iodine foods (kelp/seaweed), and two high-UIC studies used mean not median. So the exposure axis carries the domain’s binding constraint -> Measurement Error in Dietary Assessment: an area-median UIC contrast is a coarse proxy for what any individual ingests, and the crudeness runs into every OR above. (inferred from Katagiri et al., 2017)
Decision relevance
- For an iodine-sufficient person: do not over-supplement. The lower arm’s benefit is spent once replete; past sufficiency the marginal move is toward the harm arm, not more benefit. Frame iodine as a deficiency-correction lever, not a supplement-to-optimise one -> Dietary Supplements.
- Structural lever, population-level. «Monitoring of both iodine concentration in salt as well as the iodine concentration in local drinking water are essential» (Katagiri et al., 2017) — water iodine, not only salt, is a source of chronic excess; this is an environmental-monitoring decision more than an individual one.
- The iodized-salt lever cuts both ways — switching to specialty salt removes it. Because iodized table salt is a main deficiency-correction vehicle, swapping it for non-iodized sea/Himalayan salt (a popular “natural salt is healthier” move) can lower iodine intake with no BP benefit — the salt types carry identical sodium -> Sodium Intake and Blood Pressure.
- The vulnerable strata carry the decision. Antibody-positive / autoimmune / nodular / post-thyroidectomy individuals are where excess iodine most plausibly tips SCH -> OH; the general replete population’s absolute risk is low and imprecisely known.
- Supplementation in pregnancy is itself a route to excess — the two arms meet in one stratum. Prenatal iodine supplementation raises the rate of excessive iodine intake: «7.0% of women who received iodine supplements and 1.6% of women who did not had excessive iodine intake during pregnancy (average RR 4.33; 95% CI 1.24 to 15.07)» (Harding et al., 2017). So in an already-sufficient pregnant woman, routine supplementation pushes toward this page’s harm arm rather than the deficiency-correction benefit — the pregnancy-stratum decision is characterised on Iodine Supplementation in Pregnancy. Note the tolerable-upper-intake figures (IOM 1100 µg/day, EC 600 µg/day) are «not intended for iodine-deficient populations» (Harding et al., 2017), so a deficient thyroid’s safe window is narrower, not wider.
- Not a treatment page. Whether a resulting SCH should be treated is a separate, prescriber-zone decision -> Levothyroxine for Subclinical Hypothyroidism.
Limits
- Single source; gold design, low-quality evidence. SR+MA (gold by design) but the pooled studies are cross-sectional with a high risk of bias and unadjusted analyses — the tier is the design’s, the certainty is the evidence’s, and they diverge here.
- Borderline excess (300-~340 μg/L) is unestimated — the studied range starts far above the cutoff.
- Publication bias not excluded — «Funnel plots of cross-sectional studies were not asymmetrical» but «the number of papers was limited». (Katagiri et al., 2017)
- Antibody interaction excluded by design — the highest-risk stratum is signalled, not quantified.
- Children and pregnant women: too heterogeneous to pool (SCH forest plots high I²); no adult hyperthyroidism / goitre / nodule meta-analysis (too few cases, inconsistent).
- G-gaps (needs a source): a dose-response including borderline-excess intake; an antibody-stratified review; a referent-corrected or MR-instrumented adjudication of the arm magnitude.