The three-times-upper-limit convention, and where it came from

There is a complication in the reference interval that deserves more attention than it gets. The upper limit of normal for alanine aminotransferase in most clinical laboratories was derived from reference populations that included people with undiagnosed hepatic steatosis, which inflates it. Work redefining the healthy range from a population screened for viral hepatitis, alcohol and metabolic risk found substantially lower limits — around thirty units per litre for men and in the region of nineteen for women. A person whose ALT falls from 46 to 32 has moved from clearly abnormal to normal by their laboratory interval and to borderline by a stricter one.

The creatinine artefact, stated precisely

Serum creatinine is the breakdown product of creatine phosphate in skeletal muscle, produced at a rate approximately proportional to muscle mass and cleared predominantly by glomerular filtration. Estimated glomerular filtration rate is calculated from serum creatinine with adjustments for age and sex, which function as population-average proxies for muscle mass.¹

When actual muscle mass falls, creatinine production falls, serum concentration falls, and the equation reports a higher estimated filtration rate. The magnitude is not trivial: a loss of four to five kilograms of lean tissue can shift estimated filtration rate upward by several millilitres per minute per 1.73 square metres with no change in the kidney whatever. The effect runs in the reassuring direction, which is why it is rarely questioned.

The check is cystatin C, a low-molecular-weight protein produced by all nucleated cells at a rate largely independent of muscle mass. Where creatinine-based and cystatin C-based estimates diverge substantially during rapid weight loss, the divergence is itself informative, and combined equations using both are available and better validated than either alone. Cystatin C has its own confounders — corticosteroids, thyroid dysfunction and adiposity all affect it — which is why the recommendation is to read the two together rather than to substitute one for the other.

What the renal outcome trial measured

The renal outcome programme in type 2 diabetes with chronic kidney disease is the only trial in this class designed with kidney endpoints as its primary purpose. It randomised participants with established chronic kidney disease and reported a reduction in a composite of kidney disease progression, kidney death and cardiovascular death, together with a slower annual decline in estimated glomerular filtration rate, over a median follow-up of several years.²

Two features of the eGFR data matter for anybody reading a panel. There is an initial dip in estimated filtration rate on starting treatment, of the order of one millilitre per minute per 1.73 square metres, which resolves and is followed by a slower long-term decline than in the comparator arm. That pattern — an acute dip followed by long-term preservation — is familiar from other renoprotective drug classes and is generally understood as a haemodynamic effect rather than injury.

The practical implication is that a small fall in eGFR in the first months of treatment is expected and is not evidence of harm, while a large fall is not expected and is. Distinguishing them requires knowing the reference change value for creatinine, which is around fourteen per cent, and knowing whether the person has been vomiting, which changes everything.

The earlier cardiovascular outcome trials in the class carried renal composites as secondary endpoints and reported reductions in new or worsening nephropathy driven largely by albuminuria, which is a weaker endpoint than the eGFR-based composites of the dedicated renal trial.³⁴ Anybody quoting renal benefit from those programmes should say which component of which composite they mean.

The alanine aminotransferase interval is too wide

Most clinical laboratories report an upper limit of normal for alanine aminotransferase somewhere between about 40 and 55 units per litre, with a modest sex difference or none. Those intervals were derived from reference populations that were screened for viral hepatitis and heavy alcohol use but not, in most cases, for hepatic steatosis — which was neither commonly diagnosed nor considered when many of the intervals were established.

Work redefining the healthy range in a large population of prospective blood donors, screened for viral markers, alcohol intake and metabolic risk factors, arrived at substantially lower limits: in the region of 30 units per litre for men and around 19 for women.⁵ Those figures have been influential in hepatology and have largely not propagated into general laboratory reporting.

The consequence for this population is direct. A person starting treatment with an ALT of 44 has a flagged result by a strict standard and an unflagged one by their laboratory interval; a fall to 31 during treatment represents normalisation by one standard and continued abnormality by the other. Neither reading is wrong. The Journal reports ALT against both where it can, and regards a laboratory report giving only the wider interval as incomplete rather than incorrect.

Why the transaminases fall

The fall in alanine aminotransferase during successful treatment is one of the few laboratory movements in this field with a directly demonstrated mechanism, because liver fat was measured by imaging in several programmes rather than inferred from enzymes. A trial of semaglutide in biopsy-confirmed steatohepatitis reported resolution of steatohepatitis without worsening of fibrosis in a substantially greater proportion of treated participants than placebo, with corresponding falls in transaminases.⁶ The larger phase 3 programme in the same indication subsequently reported histological improvement on both resolution and fibrosis endpoints.⁷

Alongside that sits the imaging evidence from the diabetes programme, where liver fat content measured by magnetic resonance fell considerably more on a dual agonist than on insulin at broadly comparable glycaemic control, which separates the hepatic effect from the glycaemic one.

What this establishes is that the falling ALT is tracking a real change in the liver rather than reflecting reduced enzyme release for some incidental reason. What it does not establish is how much of the change is attributable to the weight loss and how much to a direct hepatic effect, since the two are not separable in a trial where the treated arm also lost more weight.

Asymptomatic pancreatic enzyme elevation

Amylase and lipase rise modestly on treatment with this drug class, by something in the region of ten to twenty per cent on average, and elevations above the upper reference limit are more common on drug than on placebo. This has been characterised most thoroughly in the liraglutide cardiovascular outcome programme, which followed more than nine thousand participants for a median of 3.8 years and therefore had the events to adjudicate.⁸ A dedicated analysis within it found higher mean enzyme concentrations on treatment with no corresponding excess of adjudicated acute pancreatitis, and concluded that the elevations had no useful predictive value for the clinical event.⁹

The diagnostic threshold for acute pancreatitis is a lipase above three times the upper reference limit in the presence of characteristic abdominal pain, or imaging evidence. Both limbs are required. A lipase of twice the upper limit in an asymptomatic person on treatment is a common finding with no established significance, and investigating it as though it were the first limb of a diagnosis produces imaging, anxiety and no information.

The Journal notes that this is one of the few places in this subject where the trial evidence is genuinely clarifying: somebody asked the question directly, measured the enzymes systematically, adjudicated the clinical events independently, and reported that the two did not track. That is what a useful safety analysis looks like.

The lipid panel: what moves, and because of what

During substantial weight loss on these agents, triglycerides fall markedly — reductions of the order of twenty per cent are reported in the obesity programmes — high-density lipoprotein cholesterol rises modestly, and low-density lipoprotein cholesterol falls only slightly.¹⁰ That pattern is the signature of weight loss and improved insulin sensitivity rather than of a lipid-lowering drug effect, and it is worth saying so, because the class is sometimes described as though it were one.

Two measurement points matter. Triglycerides have large within-person biological variation, with a reference change value above thirty per cent, so an individual fall of twenty per cent between two panels may be noise even though the group mean fall of twenty per cent in a trial is a solid finding. And fasting is no longer required for routine lipid assessment; non-fasting samples differ trivially for total and LDL cholesterol and modestly for triglycerides, and international consensus has favoured non-fasting measurement for a decade.¹¹

Lipoprotein(a) is worth a separate sentence because it is the exception. It is largely genetically determined, changes little with weight loss, and if it is going to be measured at all it needs measuring once rather than monitored. A person expecting it to improve alongside everything else will be disappointed by a result that was never going to move.

How the Journal reports a laboratory figure

Five things accompany a laboratory number in these pages. The units, because international and conventional units differ for several analytes and the same value means different things in each. The reference interval used, with a note where the interval is contested, as it is for alanine aminotransferase. The baseline, because a change of 1.8 percentage points in HbA1c from a starting value of 8.3 is a different claim from the same change from 9.5. The estimand where the figure comes from a trial. And the reference change value where we are discussing an individual delta rather than a group mean.

We also state the assay method where it matters, which is more often than one would like: HbA1c in the presence of a haemoglobin variant, thyroid function in the presence of interfering antibodies, and creatinine measured by enzymatic against Jaffe methods all behave differently, and a comparison across methods is not a comparison.

This is a heavier apparatus than most publications carry and it exists because the alternative, in our experience, is a stream of technically accurate figures that lead readers to conclusions the data does not support. Errors in this apparatus should be reported to standards@compoundjournal.com; the correction log records what came of each one.

The Journal’s position on monitoring is that the panel is the easy part and the interpretation is the hard part, and that almost all of the available effort goes into the first. A baseline panel is worth having because it converts every subsequent result from a population comparison into a personal one. Beyond that, the frequency at which most people are being tested generates flags faster than information, and there is no evidence that it detects anything a symptom-prompted panel would miss.