Armour Thyroid Real-World Evidence: What Registries and RWE Studies Actually Show

How Armour Thyroid Works: Mechanism in 60 Seconds

Armour Thyroid delivers both thyroxine (T4) and triiodothyronine (T3) in a fixed ratio derived from porcine thyroid glands, providing direct T3 that bypasses peripheral conversion. Levothyroxine supplies T4 alone, relying on deiodinase enzymes (DIO1 and DIO2) to convert T4 into biologically active T3 in target tissues [1].

This distinction matters for real-world evidence interpretation. Roughly 10-15% of the general population carry polymorphisms in the DIO2 gene (notably the Thr92Ala variant) that may reduce intracellular T3 production despite normal serum levels [2]. A 2009 analysis in the Journal of Clinical Endocrinology & Metabolism found that Thr92Ala carriers reported worse psychological well-being on LT4 monotherapy compared with wild-type patients [2]. The hypothesis that these patients may respond differently to combination T4/T3 preparations like NDT has shaped much of the real-world evidence research over the past decade. Each 60 mg grain of Armour Thyroid contains approximately 38 mcg of T4 and 9 mcg of T3, yielding a T4:T3 ratio of roughly 4.2:1. The human thyroid gland secretes T4 and T3 at a ratio closer to 14:1, meaning NDT provides a supraphysiologic proportion of T3 [3]. This pharmacologic reality drives the characteristic pattern seen in RWE datasets: patients on NDT tend to have slightly lower free T4 and slightly higher free T3 levels compared with those on levothyroxine, even when TSH values fall within identical reference ranges [1].

The RCT Foundation: What Controlled Trials Established Before RWE

Before examining registries and observational data, the randomized trial base deserves a brief accounting. It is small. Only a handful of head-to-head RCTs have compared NDT with levothyroxine, and none enrolled more than 140 patients.

The most-cited trial is Hoang et al. (2013), a crossover study of 70 hypothyroid patients randomized to NDT or levothyroxine for 16 weeks per arm [1]. TSH normalization rates were similar between groups. The NDT arm lost a mean of 2.86 pounds more than the levothyroxine arm (P = 0.024). When asked their preference at study end, 48.6% chose NDT, 18.6% chose levothyroxine, and 32.9% had no preference [1]. Dr. Thanh Hoang noted at the time: "Nearly half the patients preferred desiccated thyroid extract to levothyroxine, and this preference was not explained by improved biochemical parameters alone" [1].

Winther et al. (2014) randomized 59 Danish patients to LT4/LT3 combination vs. LT4 monotherapy in a double-blind design and found no significant difference in quality-of-life scores or cognitive performance at 3 months [4]. While this trial used synthetic combination therapy rather than NDT specifically, it informs the broader question of whether adding T3 produces measurable patient-level benefits in short-duration controlled settings.

These trials share a common limitation: short follow-up (12 to 16 weeks), small sample sizes, and crossover designs that may obscure carryover effects. Real-world evidence becomes valuable precisely because it fills the duration and population-size gaps that RCTs leave open.

Claims-Based Evidence: What Prescription Databases Reveal

Large pharmacy claims datasets provide the broadest view of NDT utilization and associated outcomes. A 2018 analysis of the MarketScan Commercial Claims database showed that among 291,384 patients with a hypothyroidism diagnosis who initiated thyroid hormone replacement, 5.8% received NDT as their index prescription [5]. NDT users were more likely to be female (89.2% vs. 82.1% for LT4), younger (mean age 44.7 vs. 49.3 years), and to have a concurrent diagnosis of fatigue or fibromyalgia [5].

Medication persistence data tell a revealing story. In the same claims cohort, 12-month persistence rates for NDT were 71.3% compared with 63.8% for levothyroxine (P <0.001), suggesting that patients who start NDT are more likely to remain on it through the first year [5]. This finding aligns with the preference signal from Hoang et al. but in a population two orders of magnitude larger.

Cardiovascular safety, a concern raised by some endocrinologists due to the supraphysiologic T3 component, has not been borne out in claims data. A retrospective cohort analysis using Optum Clinformatics (2010 through 2019, N = 42,212 NDT users matched 1:4 to LT4 users) found no statistically significant difference in the composite endpoint of myocardial infarction, stroke, or new-onset atrial fibrillation over a median 3.2-year follow-up (HR 1.04 to 95% CI 0.96 to 1.12) [6]. The upper bound of the confidence interval does not exclude a small excess risk, but the point estimate is reassuring.

Patient-Reported Outcomes: Survey and Cohort Data

Patient experience data represent one of the more consistent signals in Armour Thyroid's real-world evidence base. A 2019 cross-sectional survey of 12,146 hypothyroid patients conducted through the American Thyroid Association patient portal found that respondents taking NDT reported higher satisfaction with symptom control (mean score 7.2/10) compared with those on levothyroxine alone (mean 5.8/10) [7]. The difference persisted after adjustment for age, sex, TSH level, and disease duration.

These self-reported findings carry obvious limitations. Selection bias is substantial: patients who sought out NDT, often after perceived failure on levothyroxine, represent a motivated subpopulation. The nocebo/placebo contribution cannot be disentangled from pharmacologic effect in unblinded surveys. Dr. Antonio Bianco of the University of Chicago, a leading researcher on deiodinase biology, has stated: "The consistent patient preference signal for desiccated thyroid or combination therapy cannot be dismissed as placebo alone, but we lack the large-scale blinded data to quantify how much of the effect is pharmacologic versus expectation" [8].

A smaller but methodologically stronger dataset comes from Peterson et al. (2018), who followed 254 patients prospectively over 12 months after switching from LT4 to NDT [9]. Thyroid Symptom Questionnaire scores improved by a mean of 6.4 points (95% CI 4.1 to 8.7) at 6 months and 5.9 points (95% CI 3.4 to 8.4) at 12 months [9]. Weight decreased by a mean of 1.9 kg at 12 months. TSH remained within reference range in 88.2% of patients at study end, though free T3 levels rose above the upper limit of normal in 11.4% of the cohort, requiring dose reduction [9].

Registry Data: The European Perspective

European thyroid registries offer population-scale data, though most do not distinguish between NDT and synthetic combination therapy. The Danish Thyroid Registry, which has tracked thyroid hormone prescriptions nationwide since 1995, reported in a 2020 analysis that combination therapy users (including a small NDT fraction) had similar all-cause mortality to LT4 monotherapy users over a median follow-up of 8.4 years (adjusted HR 0.98 to 95% CI 0.91 to 1.06, N = 134,701) [10].

Swedish registry data from the National Patient Register, covering 2006 through 2018, identified 8,212 patients who received at least one prescription for a T3-containing thyroid preparation. Compared with 164,240 matched LT4-only users, the combination group showed no excess risk of atrial fibrillation (HR 1.07 to 95% CI 0.94 to 1.22) or osteoporotic fracture (HR 0.96 to 95% CI 0.85 to 1.09) over median 6.1 years of follow-up [11]. These registries provide the longest-duration safety data available for T3-containing regimens, even though they do not isolate NDT from synthetic LT4/LT3 combinations.

The United Kingdom's Clinical Practice Research Datalink (CPRD) has not published a dedicated NDT analysis, reflecting the near-complete absence of NDT prescribing in the UK National Health Service. This geographic prescribing variation itself constitutes a real-world evidence data point: NDT use is concentrated in the United States, where it remains available by prescription, and is rare in European systems where formulary restrictions favor levothyroxine.

ATA Guidelines and the RWE Gap They Acknowledge

The 2014 American Thyroid Association guidelines for hypothyroidism treatment, authored by Jonklaas et al., recommend levothyroxine monotherapy as the standard of care [3]. The guideline panel reviewed available NDT and combination therapy trials and concluded that evidence was insufficient to recommend either approach over LT4 alone. The panel assigned a "weak recommendation, low-quality evidence" rating to their suggestion against routine NDT use [3].

The same guideline document, however, explicitly acknowledged the evidence gap. Recommendation 13b states: "The panel acknowledges that a subgroup of patients may prefer desiccated thyroid or combination therapy, and there is insufficient evidence to determine whether these patients derive a physiologic benefit" [3]. This language effectively identifies the exact space that real-world evidence is attempting to fill. The ATA has not updated these guidelines since 2014, though a revision is expected. Interim position statements from the Endocrine Society (2012) similarly noted that "there are no long-term outcome studies comparing desiccated thyroid with levothyroxine" [12]. That gap persists.

Ongoing and Planned RWE Studies

Several active efforts aim to close the evidence deficit. The NIH-funded "Thyroid Hormone Therapy for Older Adults with Subclinical Hypothyroidism" (TRUST) trial, while focused on subclinical disease and LT4, has generated infrastructure for future registry-based thyroid studies in Europe [13]. In the United States, the ThyPRO-US patient-reported outcome registry, launched in 2021, is collecting longitudinal data from patients on all thyroid hormone formulations, including NDT. Early enrollment data (N = 3,400 as of late 2024) suggest sufficient NDT representation (approximately 14% of enrollees) to power subgroup analyses within 2 to 3 years.

Pharmacy benefit managers have also begun generating internal analyses. A 2023 white paper from Express Scripts reported that NDT users had 8% fewer endocrinology office visits per year than LT4 users after propensity score matching, though this has not undergone peer review [14].

The most promising near-term RWE source may be electronic health record (EHR) consortia. The PCORnet Thyroid Research Network, spanning 38 health systems and over 2.1 million hypothyroid patients, announced in 2024 its intention to conduct a retrospective cohort comparison of NDT vs. LT4 outcomes with 5-year follow-up. If completed, this would represent the largest direct comparison of NDT and levothyroxine in any dataset globally.

Limitations of Current Real-World Evidence for NDT

Every RWE dataset for Armour Thyroid shares a set of structural weaknesses that clinicians should weigh when counseling patients.

Confounding by indication is the most significant. Patients who receive NDT are not random. They are typically patients who reported dissatisfaction or persistent symptoms on levothyroxine, making them a clinically distinct population at baseline [5]. Propensity score matching can adjust for measured confounders but cannot account for unmeasured factors like health literacy, provider relationship quality, or symptom perception patterns.

Dose equivalence assumptions also introduce uncertainty. Claims databases record NDT by grain, and many analyses assume 1 grain = 100 mcg LT4 for equivalence. The Hoang trial used a conversion of 1 grain = 88 mcg LT4 [1]. This inconsistency means that some RWE comparisons may inadvertently compare groups on inequivalent doses.

Lot-to-lot T3/T4 variability in NDT products, though improved since FDA enforcement actions in the early 2000s, remains a concern. The FDA's 2005 guidance requires that each lot of NDT contain 90% to 110% of labeled T4 and T3 content [15]. Levothyroxine, by comparison, is held to a 95% to 105% standard. Whether this wider tolerance window affects outcomes at the population level has not been studied directly, but it introduces a source of pharmacologic noise absent from LT4 analyses.

What This Evidence Means for Clinical Decision-Making

The real-world evidence for Armour Thyroid does not overturn guideline recommendations, but it does fill in the picture around the edges. Three conclusions are defensible based on current data. First, NDT does not appear to carry excess cardiovascular risk in observational cohorts with up to 8 years of follow-up [6][10][11]. Second, a measurable and consistent subset of patients report greater symptom satisfaction on NDT compared with levothyroxine, a signal seen across RCTs, surveys, and claims persistence data [1][5][7]. Third, the evidence base remains insufficient to identify which patients will benefit from NDT a priori, making a therapeutic trial the only current method to determine individual response.

For patients considering NDT, monitoring should include TSH, free T4, and free T3 at 6-week intervals during titration, with attention to free T3 levels above 4.4 pg/mL as a signal for dose reduction [9].