When a better test result does not mean better health

Imagine that a treatment improves your blood results, but you still have the same symptoms and the same difficulty getting through the day. The improvement may tell us something useful about your biology. It leaves a further question: has the treatment helped your health in a way that matters?

Researchers call a measured biological feature a biomarker. Blood pressure, blood proteins and measurements from scans can all be biomarkers. They help identify risk, diagnose disease and show whether a treatment affects its intended target. Some also reliably predict clinical benefit in particular settings. Trials showing fewer strokes after blood-pressure treatment are a familiar example. [7]

A new marker needs that connection established for its own use. Three recent research programs show why.

One medicine, different health questions

The EVOKE and EVOKE+ trials tested daily oral semaglutide in 3,808 people with early symptomatic Alzheimer’s disease. After about two years, memory and everyday function declined along similar paths in the semaglutide and placebo groups. The sponsor reported improvements in some Alzheimer’s-related biomarkers, but the treatment did not slow the clinical progression the trials were designed to test. [1] [2]

A different study, the SELECT trial, found a meaningful cardiovascular benefit from weekly semaglutide injections. It enrolled people with established cardiovascular disease and overweight or obesity, without diabetes. Over roughly three years, cardiovascular death, heart attack or stroke occurred in 6.5% of the semaglutide group and 8.0% of the placebo group. More people receiving semaglutide also stopped treatment because of adverse effects. [8]

Both findings matter. The Alzheimer’s result does not erase the cardiovascular benefit, and the cardiovascular result does not establish dementia prevention. They concern different formulations, groups and health goals. Someone taking semaglutide for an established reason should discuss changes with their clinician rather than infer them from an unrelated trial.

Lowering a risk marker still needs an outcome test

Lipoprotein(a), or Lp(a), is a largely inherited contributor to cardiovascular risk. Pelacarsen was developed to lower it. The HORIZON trial tested whether that reduction would lead to fewer major cardiovascular events in 8,323 people with elevated Lp(a) and existing cardiovascular disease. [3] [4]

In September 2026, Novartis announced that the trial had missed its main cardiovascular goal despite lowering Lp(a). The blood result showed that the drug affected its target. It did not establish the hoped-for protection against clinical events.

This does not make high Lp(a) irrelevant. The National Lipid Association continued to emphasize its importance and the value of managing overall cardiovascular risk. It means that the benefit of this particular treatment must be judged by what happened to patients, alongside the change in their blood tests. [9]

Finding cancer earlier and helping people live longer

Galleri is a blood test designed to detect signals associated with several cancers. The NHS-Galleri trial examined adding it to usual screening. Its main question was whether fewer people would be diagnosed with stage III or IV cancer, considered together in a specified group of cancer types. That main goal was not met. [5] [6]

A separate analysis found fewer stage IV diagnoses. That is an encouraging finding because stage IV disease is especially serious. It still does not show that fewer people died from cancer; the trial is following that question over time. [6]

The distinction can seem surprising. Diagnosing a cancer sooner starts the clock earlier, so time lived “after diagnosis” can increase even if the date of death does not change. Screening can also find a cancer that would never have caused harm, leading to unnecessary investigation or treatment. Researchers therefore need to consider false alarms, missed cancers, follow-up procedures and deaths, as well as earlier detection. [10]

Why the original question matters

A study chooses its main outcome in advance so that researchers cannot simply search among many results and present the most favorable one as the answer. Additional findings may be valuable and can inspire the next trial. They need to be described as additional findings. [11]

At Healthy Longevity Clinic, our approach to judging treatment evidence connects each measurement with the benefit someone hopes to achieve. If the goal is preserving memory, we look for preserved memory and function. If the goal is preventing disease, we ask whether disease became less common and what harms accompanied treatment.

Bring a promising test result to that conversation. Ask what the measurement means, whether changing it predicts the benefit you want, and what effective care is available now. Better measurements help us understand the body. Their greatest value is in helping people live better.