What 200,000+ Lab Results Revealed About Nutrient Status in Integrative Practice

You can't improve what you don't measure. A first look at nutrient status across 82,403 de-identified patient profiles in the Fullscript network.

Routine laboratory testing helps answer many important clinical questions, but not all of them. When patients present with persistent fatigue, brain fog, muscle cramps, low mood, or other nonspecific symptoms, practitioners often expand their evaluation beyond routine laboratory panels to better understand whether nutritional status may be contributing to the clinical picture. Ferritin, omega-3 index, magnesium, vitamin D, and other nutrient biomarkers can provide additional context that routine screening may miss.

Until now, however, there has been little opportunity to step back and ask a broader question:

What do these tests reveal when viewed across an entire clinical network?

To explore that question, we examined a frozen dataset containing 273,844 de-identified nutrient-related laboratory results associated with 82,403 de-identified patient profiles across the Fullscript and Rupa platforms. Of these, 200,960 profile-marker results were classifiable using the current Fullscript product bands and were included in the displayed analyses. Rather than evaluating individual patient care, we looked for broader clinical patterns, such as which nutrient results most often fall outside Fullscript’s optimal bands, where Fullscript and conventional laboratory thresholds diverge, and what those patterns may reveal about laboratory interpretation in personalized, whole-person care.

For several biomarkers, results classified outside Fullscript’s optimal bands were common among the patient profiles tested. More importantly, the analysis highlights how strongly the clinical picture can depend on the biomarker measured and the thresholds used to interpret the result.

At a glance

Across 200,960 classifiable profile-marker results from 82,403 de-identified patient profiles, three findings stood out:

• Omega-3 status (EPA+DPA+DHA) was below the Fullscript optimal floor in 93% of displayed profile results, while profiles with this test represented only 6.3% of the nutrient-test cohort.
• Among female-profile ferritin results, 48.5% fell below the Fullscript optimal floor, while 34.8% fell between Quest’s conventional lower reference limit and the Fullscript optimal floor, illustrating how interpretation can change depending on the threshold applied.
• RBC magnesium provided one of the clearest examples of threshold sensitivity: 88.4% of displayed results fell below Fullscript’s 6.0 mg/dL optimal floor, compared with 4.4% below Quest’s 4.0 mg/dL lower reference limit.

These findings describe tested patient profiles rather than the general population, but they provide a useful snapshot of real-world testing patterns and how interpretation can change depending on the biomarker and classification thresholds used.

How we looked

This analysis examined a frozen cohort of 273,844 unit-eligible nutrient-related laboratory results collected from January 1, 2023 through March 31, 2026, associated with 82,403 de-identified patient profiles across the Fullscript and Rupa platforms. Of these, 200,960 profile-marker results were classifiable using the current Fullscript product bands and included in the displayed analyses.

For each biomarker, only the latest eligible result per patient profile was retained to avoid giving profiles with repeat testing additional weight. These findings therefore represent a cross-sectional snapshot of the latest available measurement rather than baseline nutrient status; repeat-testing patterns and treatment effects were not evaluated. Future analyses may examine longitudinal changes among profiles with repeat testing.

Note: Results were classified using Fullscript’s currently-defined ‘optimal ranges’, as established through evidence review by Fullscript’s Medical Advisory team. These bands are often tighter than conventional laboratory reference intervals, so a result classified as suboptimal by Fullscript may not be flagged on a standard laboratory report. Where available, published Quest ranges or interpretive categories are used as a consistent contextual comparator rather than a universal conventional standard. Optimal ranges are provided solely for informational reference content. Optimal ranges are not medical diagnoses or treatments, are not a substitute for a practitioner's professional judgment in specific individual situations, and are not meant to provide medical or professional advice.

Nutrient biomarker band distributions

The figure below provides an overview of all ten nutrient biomarkers, ranked by the proportion of tested patients who fell below the Fullscript optimal range. Together with testing frequency, it provides context for the findings explored below.

Finding 1: Omega-3 status

Infrequently measured, frequently below the Fullscript optimal floor.

Among all ten nutrient biomarkers included in this analysis, the omega-3 index stood out immediately. Of 5,209 displayed EPA+DPA+DHA profile results, 93.2% fell below the Fullscript optimal floor of 8% by weight, giving it the largest below-optimal share in the analysis.

Profiles with an EPA+DPA+DHA result represented 6.3% of the nutrient-test cohort, making it one of the least frequently represented biomarkers in the analysis. The reasons practitioners choose to order omega-3 testing cannot be determined from these data; the pattern could reflect targeted testing, routine supplementation without laboratory confirmation, or other clinical considerations.

This raises an important clinical question: What is the appropriate role of omega-3 testing in clinical practice?

This analysis cannot answer that question directly. Patients who receive omega-3 testing may not be representative of the broader population receiving care, and selection bias likely contributes to the high prevalence observed. The below-optimal proportions were descriptively similar across the Fullscript and Rupa sources when the same Fullscript cut-points were applied. These source-specific rates provide context rather than independent replication, because the sources differ in era, patient mix, practitioner mix, data ingestion, and potentially assay methodology.

Why it matters

Omega-3 status is increasingly recognized as an important component of personalized, whole-person care, yet it remains one of the least frequently measured nutrient biomarkers in this dataset. The combination of high below-optimal share and relatively low testing frequency makes this one of the most notable findings in the analysis. Rather than suggesting that every patient should be tested, it highlights an opportunity to better understand when omega-3 testing provides meaningful value alongside clinical judgment.

93.2% of displayed EPA+DPA+DHA results fell below the Fullscript optimal floor, while profiles with this test represented 6.3% of the nutrient-test cohort.

Finding 2: Ferritin

The gap between ‘normal’ and ‘optimal’

If the omega-3 findings highlight an opportunity to expand testing, ferritin highlights one of the most common interpretation challenges in personalized care: what happens when a result is technically ‘normal’ but may not be optimal for an individual patient?

Among 11,536 displayed female-profile ferritin results, 48.5% fell below the Fullscript optimal floor of 40 ng/mL. Of all displayed female-profile results, 34.8% fell between Quest’s 16 ng/mL lower reference limit and Fullscript’s 40 ng/mL optimal floor. This threshold-comparison group represented 71.8% of the female results classified below Fullscript optimal. Because the analysis also includes age<18 and unknown-age profiles, this should be interpreted as a threshold comparison rather than a claim that every result fell within its age-applicable Quest reference interval.

This is where clinical context matters: a result may sit above a conventional lower limit yet still deserve a closer look when the patient’s symptoms, history, or complementary findings raise concern for iron insufficiency.

This should not be interpreted as a zone of missed disease. It is a zone where clinical interpretation may differ. The data quantify how many results fall within this threshold-comparison range, but they do not determine which threshold should guide care or whether intervention is warranted.

Ferritin tells a different story in males

While the primary finding centered on females with below-optimal ferritin, the distribution among male patient profiles was almost the mirror image.

Unlike low ferritin, however, higher ferritin results are rarely interpreted in isolation. Ferritin is an acute-phase reactant, and higher concentrations may reflect inflammation, metabolic dysfunction, liver disease, iron overload, or other underlying clinical processes.

The upper-band pattern in this analysis should therefore be interpreted cautiously. The analysis does not include inflammatory, liver, metabolic, or other clinical context needed to determine why an individual ferritin result was elevated. Among 2,634 displayed male-profile ferritin results, 69.9% fell within the combined Fullscript Suboptimal High or High bands, but the analysis does not include the inflammatory, liver, metabolic, or other clinical context needed to determine why an individual result was elevated. These findings should be interpreted within the broader clinical picture rather than as a standalone conclusion.

Why it matters

Ferritin was one of the clearest examples in this analysis of how laboratory interpretation can influence clinical decision-making. Whether evaluating a patient with fatigue, reduced exercise tolerance, hair loss, heavy menstrual bleeding, or other symptoms suggestive of iron deficiency, or investigating unexpectedly elevated ferritin that may warrant further assessment, the clinical significance of the result depends on more than the laboratory value alone.

Together, these findings reinforce that ferritin is more than a marker of iron status. Its clinical value lies in thoughtful interpretation, and in considering the patient's symptoms, history, complementary biomarkers, and the clinical question being asked, rather than relying on a single laboratory threshold alone.

34.8% of displayed female-profile ferritin results fell between Quest’s 16 ng/mL lower reference limit and Fullscript’s 40 ng/mL optimal floor.

Finding 3: Magnesium

Why what you measure and how you interpret it both matter.

Magnesium provides one of the clearest examples in this analysis of how strongly laboratory interpretation can depend on the framework applied. Among 10,638 displayed RBC-magnesium profile results, 88.4% fell below Fullscript’s optimal floor of 6.0 mg/dL. By comparison, 4.4% fell below Quest’s lower reference limit of 4.0 mg/dL.

Most of that difference lies between the two thresholds: 83.9% of displayed RBC-magnesium results were at or above Quest’s lower reference limit but below Fullscript’s optimal floor. In other words, the same measured distribution can tell a very different story depending on the cut-point used. This analysis does not establish which threshold should guide individual patient care.

But threshold choice is only part of the magnesium story. The test itself also determines what aspect of magnesium status is being measured. Serum magnesium reflects a tightly regulated extracellular pool and remains the most widely used and practical clinical measure. Because only a small proportion of total-body magnesium circulates extracellularly, however, a normal serum value does not necessarily reflect intracellular or total-body magnesium status.

RBC magnesium measures magnesium within erythrocytes and therefore offers an intracellular perspective. Some research suggests that RBC magnesium can provide information not captured by serum testing, although no single biomarker has been shown to reliably represent whole-body magnesium status in every clinical context (Jahnen-Dechent 2012, Rude 1998, Fiorentini 2021, Xiong 2019).

Importantly, the serum and RBC results in this analysis came from largely different patient-profile subsets rather than paired measurements in the same individuals. The 1.5% below-optimal share for serum magnesium and 88.4% share for RBC magnesium therefore should not be interpreted as evidence that RBC testing detects magnesium depletion more effectively. Differences may reflect the thresholds applied, patient selection, laboratory source, assay methodology, physiology, or ordering practices.

Why it matters

For providers, magnesium reinforces a broader principle of personalized laboratory medicine: the result only has meaning in the context of what was measured, how it was classified, and why the test was ordered. Routine serum testing can answer important questions about circulating magnesium, while alternative measures may provide complementary information when the clinical question extends beyond the extracellular compartment.

A whole-person approach therefore does not mean automatically ordering more testing or assuming that one assay is superior. It means understanding the strengths and limitations of the available measures and interpreting them alongside the patient’s symptoms, history, diet, medications, comorbidities, and broader clinical picture.

88.4% of displayed RBC-magnesium results fell below Fullscript’s 6.0 mg/dL optimal floor, compared with 4.4% below Quest’s 4.0 mg/dL lower reference limit, a reminder that the boundary used can dramatically change how the same measurements are classified.

Other observations

While omega-3 status, ferritin, and magnesium emerged as the strongest clinical stories, several additional observations help round out the broader picture of nutrient testing in everyday practice.

Vitamin D remained the most frequently represented nutrient biomarker, with results available for 81.7% of the nutrient-test cohort. Of displayed vitamin D results, 24.1% fell below the Fullscript optimal floor of 30 ng/mL.

Serum folate also had a high below-optimal share, with 58.6% of displayed profile results falling below Fullscript’s optimal threshold of 14.8 ng/mL.  Serum folate is sensitive to  recent dietary intake, however, and should be interpreted within the broader clinical context.

Vitamin B12 provided another example of threshold sensitivity. Using Fullscript’s current optimal floor of 500 pg/mL, 35.8% of displayed profile results fell below optimal, compared with 0.4% below Quest’s lower reference limit of 200 pg/mL. These figures represent different classifications of the same measured values and do not establish which threshold should guide care, although some evidence suggests an ‘optimal’ range sits closer to 400–700 pg/mL (Krzywański 2020, Tang 2024).

Zinc and copper showed a similar threshold-dependent pattern. Among displayed results, 24.5% of zinc and 29.8% of copper results fell below Fullscript’s optimal floors, compared with 5.5% and 3.8%, respectively, below the Quest lower reference limits. Both biomarkers were less frequently represented in the dataset, and limitations in specimen-level mapping mean these findings are best viewed as descriptive rather than as evidence of a specific clinical subgroup effect.

Exploratory analyses also examined whether the observed patterns varied by provider modality or geographic region. No consistent differences emerged that materially changed the overall findings, although geographic analyses were limited to profiles with available regional information.

Why this research matters

The goal of this analysis wasn't to determine how providers should diagnose or treat nutrient deficiencies. Instead, it was to better understand what providers are seeing when they choose to look beyond routine laboratory testing.

Taken together, three themes emerged.

First, results classified below Fullscript’s optimal bands were common for several biomarkers among the patient profiles tested. These findings describe a provider-selected tested cohort and should not be interpreted as population prevalence or diagnoses of nutrient deficiency.

Second, how a result is classified matters. Ferritin, RBC magnesium, and vitamin B12 each demonstrate how the same underlying measurements can produce meaningfully different classifications depending on the thresholds applied.

Finally, personalized care isn’t simply about ordering more tests. It’s about choosing tests thoughtfully, understanding how their results are classified, and integrating those findings with the patient’s symptoms, history, and goals. Omega-3 status, ferritin, and magnesium each highlight a different part of that clinical process.

For years, conversations about nutrient status have relied largely on published research, population datasets, and clinical experience. This analysis offers something different: a first look at what these biomarkers reveal across more than 82,000 de-identified patient profiles tested through the Fullscript network. It transforms individual clinical observations into real-world clinical intelligence at scale, helping to contextualize clinical observations, identify emerging patterns, and generate new questions for future research. Rather than replacing clinical judgment, these insights are intended to strengthen it by helping providers recognize broader clinical patterns while continuing to individualize care for the patient in front of them.

How to interpret these findings

These findings are intended to provide insight into patterns observed across a large, real-world clinical dataset, not to establish prevalence in the general population or guide care for any individual patient.

When interpreting these results, it's important to keep several considerations in mind:

• These are tested patients, not the general population. Every biomarker reflects a patient profile for which a provider ordered a specific test. Clinical indication for testing was not captured prior to our analyses, so the observed distributions describe provider-selected tested subsets and should not be extrapolated to the wider population.
• Fullscript product bands and conventional laboratory comparators answer different questions. A result classified as below optimal by Fullscript may remain within a conventional laboratory reference interval. Where shown, published Quest ranges or interpretive categories are used as a consistent contextual comparator rather than a universal standard.
• These findings are observational. They identify patterns and generate hypotheses but do not establish cause-and-effect relationships or treatment recommendations.
• Clinical context remains essential. Laboratory results should always be interpreted alongside a patient's history, symptoms, examination findings, and overall clinical picture.

Study notes

• Dataset: 273,844 unit-eligible nutrient-related results across 82,403 de-identified patient profiles, resulting in 200,960 profile-marker results. Note: ‘Patient profile’ is defined as a patient-ordering account; in rare cases (estimated <0.25% of all patient profiles), a unique patient user may have more than one patient profile account. So not all results necessarily represent unique individual patients; however for the purposes of this analysis we consider them effectively interchangeable.
• Analysis time period: January 1, 2023–March 31, 2026.
• Analysis: One most recent eligible result per patient profile and biomarker. Unless otherwise noted, analyses combine Fullscript and Rupa laboratory records. Ferritin and iron analyses use Fullscript records only because their current product bands require demographic information (sex for ferritin, and sex and age for iron) that were not unavailable in the Rupa data.
• Interpretation: Results classified using Fullscript optimal ranges, with published Quest ranges or interpretive categories shown as contextual comparators where available. Adult Fullscript bands and adult Quest comparators were used as a practical convention; because age is unavailable for all Rupa records and some Fullscript records, some younger profiles may be present. Analyses requiring demographic-specific bands were limited to records with the necessary demographic information.
• Study design: Retrospective, descriptive analysis of de-identified clinical data.
• Important limitation: Findings describe provider-selected tested profiles rather than the general population and do not establish diagnoses, treatment effects, or causal relationships.

Learn more

Clinical Insights Pulse is an ongoing series exploring real-world clinical patterns across the Fullscript network. By combining large-scale de-identified clinical data with evidence-informed interpretation, the series aims to help providers better understand the patterns emerging in everyday practice, and the questions that deserve a closer look.

Labs on Fullscript classify results using clinically informed optimal-range bands alongside laboratory reference information, equipping providers with an additional framework for interpreting laboratory results within the context of personalized care.

Learn more about Labs on Fullscript and explore how clinically informed laboratory testing can support personalized patient care.

About the data behind Clinical Insights Pulse

Clinical Insights Pulse is built using HIPAA Safe Harbor de-identified clinical data from the Fullscript platform. Before any analysis begins, direct identifiers are removed and the dataset is transformed using a standardized de-identification process that has been reviewed with our Legal and Compliance teams. Researchers analyze only the resulting frozen de-identified dataset, not the original identifiable records.

All findings are designed to describe aggregated clinical patterns among tested patient profiles, not individual outcomes. Every publication undergoes Legal and Compliance review before release to help ensure privacy is protected throughout the process.

These analyses are conducted in accordance with Fullscript's applicable Terms of Service, Labs Terms of Service, and Privacy Statement, which describe how de-identified information may be used to improve our platform, advance research, and generate clinical insights.