The decision this changes
Subclinical hypothyroidism (SCH) — mildly elevated thyrotropin (TSH) with a normal free thyroxine — is common (NHANES III: ≈13 million in the US, higher in women and older people) and levothyroxine is a default prescription for it, especially when tiredness, constipation or weight gain are attributed to it. A gold meta-analysis says: in nonpregnant adults, treating SCH does not improve quality of life or symptoms. This is a de-escalation decision — a standing drug not worth starting for these outcomes in this stratum, and a lever to stop treating as a big rock. It sits in the pharmacotherapy taper (standing-drug efficacy/limitations appraisal), not the prescriber zone: the finding is stratum-level (start-or-not for the general SCH adult), not per-person titration. (inferred from Feller et al., 2018)
The effect estimate (Feller 2018 — 21 RCTs, 2192 nonpregnant adults)
SMDs coded so positive = benefit of levothyroxine (0.2/0.5/0.8 = small/moderate/large). Primary outcomes, gold SR+MA:
- «thyroid hormone therapy … was not associated with benefit regarding general quality of life (n = 796; SMD, −0.11; 95% CI, −0.25 to 0.03; I2=66.7%) or thyroid-related symptoms (n = 858; SMD, 0.01; 95% CI, −0.12 to 0.14; I2=0.0%). Overall, risk of bias was low and the quality of evidence assessed with the GRADE tool was judged moderate to high.» (Feller et al., 2018)
- Certainty was HIGH for both primary outcomes (and for muscle strength, blood pressure, BMI). The QoL SMD back-transforms to ≈0.02 on the EQ-5D and the symptom SMD to ≈0.18 on the ThyPRO (0-100) score — clinically nil. (Feller et al., 2018)
- Every secondary patient-important outcome was null too: depressive symptoms (SMD −0.10, −0.34 to 0.13), cognitive function (0.09, −0.05 to 0.22), muscle strength (0.1, −0.1 to 0.2), systolic BP (−0.7 mm Hg, −2.6 to 1.2), BMI (0.2, −0.4 to 0.8). No study measured pain. (Feller et al., 2018)
- Verdict: «These findings do not support the routine use of thyroid hormone therapy in adults with subclinical hypothyroidism.» (Feller et al., 2018)
No-meaningful-effect, not insufficient-evidence — and where the line falls
The four evidence states must be kept apart, and this source splits cleanly across two of them:
- No-meaningful-effect for QoL, symptoms, mood, cognition, muscle strength, BP, BMI: GRADE moderate-to-high, narrow CIs tight around a point estimate of ~0, and the two largest and most recent trials (incl. TRUST/Stott 2017) drive the pool. This is a positive null, not an absence.
- Insufficient-evidence for cardiovascular events and mortality: GRADE low, evaluated by a
single trial only — «definitive evidence is lacking regarding the association of therapy for
subclinical hypothy- roidism with reduced cardiovascular event rates.»
(Feller et al., 2018)
type-G
Conflating the two would either overclaim a proven-harmless CV profile or discount the genuinely strong QoL/symptom null. (inferred from Feller et al., 2018)
The surrogate the drug hits — TSH normalizes while the person does not move
Levothyroxine reliably drives the treated arm’s TSH into range (0.5-3.7 mIU/L) while placebo stays elevated (4.6-14.7 mIU/L) — «indicating that treatment was associated with nor- malization of thyrotropin levels» — yet no patient-important outcome follows. (Feller et al., 2018) This is a textbook marker-moved-outcome-didn’t case, sharpened because the surrogate (TSH) is also the diagnostic criterion: treating-to-target normalizes the number that defines the disease without changing how the person feels or functions. -> Surrogate Outcomes. (inferred from Feller et al., 2018)
Transportability — where the null may NOT hold (name the boundary)
The null is established for mildly elevated TSH in adults up to ~74 y with mild-to-moderate symptoms. It does not automatically transport to three strata, each a route-(b) effect-modification candidate the trials could not test:
- TSH >10 mIU/L (more severe SCH): «only 2 RCTs examined participants with a mean baseline thyro- tropin level higher than 10 mIU/L. Therefore, the current findings may not be generalizable to people with subclinical hypothyroidism and a thyrotropin level higher than 10 mIU/L» (baseline TSH <7.0 in 11 of 21 RCTs). (Feller et al., 2018)
- Age >80: highest mean age was 74 y, so results «may not be gener- alizable to people older than 80 years.» (Feller et al., 2018)
- High symptom burden: «It is possible that the subgroup of people with subclini- cal
hypothyroidism and a high burden of symptoms would still benefit from treatment» — severe-symptom
patients are plausibly underrepresented (treated immediately, not enrolled).
(Feller et al., 2018)
type-G
These are stated candidates, not demonstrated benefits: no positive effect-modification evidence exists yet, so under symmetric standards they stay open gaps, not carve-outs that rescue treatment. (inferred from Feller et al., 2018)
Guidance dissonance + the overtreatment cost
The MA runs against the practical reach of guidelines: «more than 90% of persons with subclinical hypothyroidism and a thyrotropin level of less than 10 mIU/L would actually qualify for treatment. How- ever, results of this meta-analysis are not consistent with these guideline recommendations. In addition to absence of an as- sociation of thyroid hormone therapy with improved out- comes, thyroid hormone therapy is associated with adverse ef- fects when overtreatment occurs.» (Feller et al., 2018) So the net at the margin is a drug that carries overtreatment harm (iatrogenic hyperthyroidism) with no offsetting patient-important benefit for the modal SCH adult — the case for not pulling this lever, and for questioning a standing prescription. (inferred from Feller et al., 2018)
Limits
- Single gold MA (loop open). Confidence in the no-benefit finding is high on this source’s own GRADE, but it is one meta-analysis; the wiki holds no independent SR to corroborate or contest it, and grades coherence, not validity. A second, independent gold SR/MA on levothyroxine for SCH (or the individual large RCTs — TRUST/Stott 2017, Zhao 2016) would upgrade the independent-backing state; none is held yet.
- Scope is treatment of established SCH for QoL/symptoms; it says nothing about frank hypothyroidism, pregnancy (excluded — the correct general-adult stratum), or SCH-plus-comorbidity populations (excluded as unrepresentative). (inferred from Feller et al., 2018)