Reference ranges are statistical, not biological. They describe where ninety-five per cent of the sample population sits — the central band — and flag the outer 2.5 per cent on each tail. A result inside the band is ‘normal.’ That is appropriate language for disease detection. It is the wrong language for performance.

If you arrive with fatigue and your full blood count, electrolytes, and TSH all come back inside their bands, three things may have happened: the right markers were measured but read against ranges that allow function to fall well below your own baseline; the right systems were not measured at all; or both. The fatigue is real. The lab report is also accurate. They are answering different questions. The longer explanation of why ‘normal’ and ‘optimal’ diverge sits in the companion piece on bloods normal but feel terrible. This article maps the specific axes that explain the fatigue presentation.

Ferritin is the storage form of iron and the most useful single marker of iron status. In Australia, the standard reference range for ferritin starts around 15–30 µg/L. Functional medicine clinicians work to a much higher floor: ferritin under 50 µg/L is treated as iron-deficient in a fatigue presentation, even when haemoglobin is in range.

The evidence backs the narrower window. A 2017 meta-analysis of randomised controlled trials in British Journal of Nutrition found that iron repletion significantly improved fatigue scores in non-anaemic women with low ferritin. The literature calls this iron deficiency without anaemia.¹ A 2022 randomised trial of intravenous versus oral iron in women with persistent ferritin under 30 µg/L found IV iron produced faster and larger improvements in patient-reported fatigue and quality of life within three to six weeks.²

The clinical implication is concrete. A woman who reports persistent fatigue, has a ferritin of 22 µg/L, and is told her bloods are normal has had iron measured correctly and interpreted against the wrong range. The trial evidence says repleting iron in that range improves how she feels. Not in months, in weeks.

The standard Australian thyroid screen is TSH alone, with Free T4 added if TSH is out of range. That misses three signals that matter for fatigue.

The first is Free T3, the active hormone. T4 is the storage form; T3 does the metabolic work. A patient can have a normal TSH and Free T4 with a low Free T3, often the result of poor T4-to-T3 conversion under chronic stress or inflammation. The second is thyroid antibodies (TPO and TgAb), which can be elevated for years before TSH moves. Antibody-positive Hashimoto’s is the most common cause of hypothyroidism in Australia and presents as fatigue, cold intolerance, and brain fog well before the lab calls it. The third is the high-normal Free T4 trap. A large Western Australian cohort study of older men (the Health In Men Study) found that high-normal Free T4 levels independently predicted frailty and fatigue, even within the euthyroid range.³ Inside the reference band, the upper quartile carried a 32–36 per cent higher odds of frailty than the lower quartile.

The full thyroid workup (TSH, Free T4, Free T3, reverse T3, TPO, TgAb) costs more than the general screen, but it also resolves more than three quarters of unexplained fatigue presentations with a thyroid component that a TSH-only test would have missed.

A single morning serum cortisol gives one point in time. The clinically useful signal is the shape of the curve across the day. Cortisol should peak about thirty minutes after waking, decline through the afternoon, and reach a low point at night.

Dysregulation shows up in three patterns: a flat curve (low energy all day), a reversed curve (tired in the morning, wired at night), and a blunted Cortisol Awakening Response (CAR), a failure to mount the morning rise. The 2016 international expert consensus on CAR measurement set the protocol used in research: salivary samples at waking, +15, +30, and +45 minutes, with strict sampling-time control.⁴ None of these patterns are visible on a single AM blood draw.

The clinical signal matters. A 2015 study in Brain, Behavior, and Immunity following 265 women undergoing cancer therapy found that physical fatigue was specifically associated with elevated evening cortisol and a flattened diurnal decline, not with morning levels or with depressive symptoms.⁵ The pattern, not the level, predicted the fatigue. A four-point salivary collection (DUTCH Adrenal or equivalent) maps overall output, and DUTCH CAR maps the awakening response. Together they give the complete picture that a single timestamp cannot.

Most metabolic screening in Australia tests fasting glucose, HbA1c, and sometimes a lipid panel. Fasting insulin is rarely included. That matters because insulin resistance is the upstream driver and it moves first.

The Whitehall II study, published in The Lancet in 2009, followed 6,538 British civil servants for nearly ten years and identified 505 cases of type 2 diabetes. The researchers then mapped backwards to characterise the metabolic trajectory: insulin sensitivity declined steeply five years before diagnosis, and fasting glucose only began its rapid rise three years before diagnosis.⁶ By the time the standard markers move, the underlying biology has been drifting for half a decade.

Brain function tracks insulin status closely. A 2010 study in Archives of Neurology measured HOMA-IR (a calculation that combines fasting glucose and insulin) and brain glucose uptake on PET imaging in cognitively normal older adults, and found higher insulin resistance correlated with Alzheimer’s-pattern reductions in cerebral glucose metabolism years before any cognitive symptoms.⁷ The fatigue presentation of early insulin resistance (afternoon energy crashes, post-meal sleepiness, slow recovery from exercise) tracks the biology directly. A fasting insulin and HOMA-IR added to a standard metabolic panel costs a few dollars and changes the window by years.

Standard testing measures total serum vitamin B12. Most of that pool is bound to haptocorrin and is not biologically active. The metabolically active fraction (the part the cell can actually use) is bound to transcobalamin and measured as holotranscobalamin (holoTC) or ‘active B12.’

A 2007 study in Clinical Chemistry of more than 2,400 older adults compared holoTC against total B12 for detecting metabolic B12 deficiency, using elevated methylmalonic acid as the gold-standard marker of cellular B12 status. HoloTC had significantly better diagnostic accuracy than serum total B12, including in patients with normal renal function.⁸ A 2012 case-based review in Nederlands Tijdschrift voor Geneeskunde described two patients with normal total B12 but clinically meaningful deficiency identified through active B12 and homocysteine.⁹

The practical version: a serum B12 of 250 pmol/L is inside the standard range. It is also the range where a meaningful subset of patients are functionally deficient. Active B12 and homocysteine, used together, separate the truly replete from the laboratory-replete. Both are available through Australian community pathology on request.