An online diabetes test built around seven measurements you can take at home drew wide coverage at the end of July 2026, after Danish researchers published a screening tool that needs no needle and no laboratory visit. The model, called MEDWACS, was developed at the Technical University of Denmark and reported in the Journal of Clinical Epidemiology. Its most unexpected input is the length of your thigh. Behind the novelty sits a very practical problem: type 2 diabetes is straightforward to diagnose once someone actually gets tested, but a large share of people never think to ask. In this article you will learn what the new tool measures, why thigh length carries information about blood sugar, where a risk estimate stops and a laboratory diagnosis begins, and what to do if a questionnaire tells you your risk is high.
What the online diabetes test actually measures
The research team started from three decades of health data collected by the US National Health and Nutrition Examination Survey, then screened close to 3,700 candidate variables. Seven survived. All of them can be captured at home with a tape measure, a bathroom scale and a standard blood pressure monitor.
- Ikä
- Seksi
- Body mass index
- Waist circumference
- Upper arm circumference
- Upper leg (thigh) length
- Systolic blood pressure
Notice what is missing: no blood sample, no fasting, no appointment. That is the whole point. As the tool’s lead author put it, type 2 diabetes is easy to diagnose with a blood test, but if you do not suspect you have it, you never go to the doctor to have one done. The tool was validated on separate datasets from the United States and South Korea, where it performed roughly on par with the risk questionnaires already used in clinical practice. It has not yet been checked against a Danish population.
Why thigh length ended up in the model
The thigh measurement surprised the researchers themselves, and there are two plausible explanations. The first is developmental: the length of the femur reflects nutrition during the first years of life, and poor early nutrition slightly restrains bone growth while raising diabetes risk decades later. The second is mechanical. The thighs hold the largest muscle group in the body, and muscle is where most circulating sugar is taken up after a meal. Less muscle mass in the legs means less capacity to clear sugar from the bloodstream, which is one of the mechanisms behind insulin resistance.
This is an association, not a cause. A short femur does not make anyone diabetic. It simply happens to summarise, in a single tape-measure reading, several things that would otherwise take a long questionnaire to capture. Readers who want the underlying physiology can consult our guide covering the insulin blood test and what fasting levels mean.
A risk score is not a diagnosis
This is the line that matters most, and it is where consumer confusion tends to start. A questionnaire sorts people into “worth testing now” and “probably not yet”. A laboratory measurement tells you where your blood sugar actually sits. The two answer different questions.
| Online risk score | Laboratory blood test |
|---|---|
| Estimates a probability from body measurements and history | Measures glucose or glycated hemoglobin directly |
| Free, instant, repeatable at home | Requires a blood draw and a certified laboratory |
| Cannot confirm or rule out diabetes | Confirms prediabetes or diabetes against defined thresholds |
| Useful to decide whether to book a test | Useful to decide what happens next clinically |
The thresholds themselves are fixed and public. According to the CDC, an A1C below 5.7% is normal, 5.7% to 6.4% indicates prediabetes and 6.5% or above indicates diabetes. The NIDDK adds that a fasting plasma glucose of 100 to 125 mg/dL and a two-hour glucose challenge of 140 to 199 mg/dL both fall in the prediabetes band. No home questionnaire produces a number on those scales. If you have results in hand, our article explains fasting blood glucose levels and their interpretation, and a companion piece covers normal glucose levels and what abnormal results mean.
Uusimmat tieteelliset edistysaskeleet
Non-invasive risk scoring is not new, and the last three years have quietly clarified what it can and cannot deliver. Here is what the recent work says, in plain terms.
A 2023 Icelandic study in primary care combined the widely used Finnish Diabetes Risk Score with an HbA1c measurement in 220 adults. The questionnaire picked up people with prediabetes reasonably well once a suitable cut-off was applied. What it could not do was replace the blood test — the actual prediabetes cases were only confirmed by the HbA1c result. What this means for you: a score points a finger, the laboratory answers the question.
Two further studies tested the same questionnaire outside Europe. One, published in 2023 from a Kenyan national survey of just over 4,000 adults, found it worked acceptably for spotting undiagnosed diabetes and remained useful in settings where laboratory capacity is limited. A 2025 analysis of a large Iranian cohort — a group of people followed over time — found similar overall accuracy but a clear difference between men and women, and recommended that risk tools be tuned for sex rather than applied identically. What this means for you: these tools travel across populations, but they are not equally accurate for everyone, which is another reason a positive score should end at a laboratory rather than at a conclusion.
The 2026 MEDWACS work fits this pattern rather than overturning it. Its contribution is convenience — seven measurements, no fasting, no clinic — and a fresh predictor nobody had used before. Its accuracy sits alongside the existing questionnaires rather than far above them, and the authors are explicit that validation in further populations is still needed. Results in this field remain preliminary until they are reproduced in the group of people who will actually use the tool.
When to ask your doctor for a blood test
A high score on an online questionnaire is a prompt, not an emergency. Bring it to a consultation and ask for glucose testing if any of the following apply.
- You are 45 or older and have never had your blood sugar checked
- You are younger but carry excess weight along with another risk factor, such as high blood pressure, a first-degree relative with diabetes or a history of gestational diabetes
- A previous result sat in the prediabetes band and was never rechecked
- You have persistent thirst, frequent urination, unexplained weight loss, blurred vision or slow-healing wounds
Seek care the same day for nausea and vomiting, deep or rapid breathing, fruity-smelling breath, confusion or abdominal pain. These can signal diabetic ketoacidosis and need immediate treatment. Before booking, it is worth reviewing verikokeen paastoamissäännöt, since a fasting glucose result is only valid if the fast was respected. When the report arrives, our guide explains miten verikokeiden tuloksia luetaan, ja erillinen artikkeli käsittelee the diabetes blood tests that confirm a diagnosis. If your doctor also looks at how hard your pancreas is working, we explain HOMA-IR-insuliiniresistenssipistemäärä and we have covered insulin resistance in people who are not overweight.
Sanasto
| Termi | Määritelmä |
|---|---|
| HbA1c (A1C) | The share of hemoglobin coated with sugar, reflecting average blood glucose over roughly three months. |
| Paastoverensokeri | Blood sugar measured after at least eight hours without food, used to screen for and diagnose diabetes. |
| Prediabetes | Blood sugar above the normal range but below the diabetes threshold, often reversible. |
| Risk score | A number estimating the probability of a condition from personal characteristics, not a measurement of the condition itself. |
| Seulonta | Testing people without symptoms to find a condition earlier than it would otherwise appear. |
| External validation | Checking a prediction tool on a separate group of people from the one used to build it. |
| Insuliiniresistenssi | A state in which cells respond poorly to insulin, so the body must produce more to keep sugar in check. |
| FINDRISC | The Finnish Diabetes Risk Score, a short questionnaire widely used to estimate ten-year type 2 diabetes risk. |
Usein kysytyt kysymykset
Can an online diabetes test diagnose diabetes?
No. Every recognised diagnostic threshold is defined on a laboratory measurement — A1C, fasting plasma glucose or an oral glucose tolerance test. A questionnaire estimates how likely you are to have a problem; it never measures your blood sugar. A high score is a reason to arrange testing, and a low score does not guarantee that your glucose is normal, particularly if you have symptoms.
Is a free online diabetes test worth taking?
It can be, if you treat the output as a prompt. Validated questionnaires such as FINDRISC are free, take a couple of minutes and are offered by several national health bodies. Their value lies in reaching people who would otherwise never consider testing. Their limit is that they were designed for populations, so an individual result carries real uncertainty. Avoid tools that ask for payment, promise a diagnosis or do not say which score they are based on.
How accurate are these home diabetes tools?
Good enough to sort people, not good enough to label them. Across studies in Europe, Africa and Asia, established risk questionnaires correctly separate higher-risk from lower-risk people most of the time, but they miss some cases and flag others who turn out to be fine. Accuracy also varies between men and women and between populations. That is exactly why every study in this field ends with the same recommendation: confirm with a blood test.
Does a normal weight mean I do not need testing?
Not necessarily. Body mass index is only one input among several, and blood sugar problems appear in people of normal weight, particularly where muscle mass is low or there is a strong family history. Age, waist circumference, blood pressure and family history all carry weight independently. If several apply to you, ask about testing regardless of what the scale says.
How often should I be screened if my result is normal?
Usual practice is to repeat glucose testing every one to three years for adults with risk factors, and annually for anyone whose previous result fell in the prediabetes range. Your own interval should be set with a clinician who knows your history, because it depends on age, family history, weight, blood pressure and previous results rather than on any single figure.
Do I need to fast before a diabetes screening test?
It depends on the test. A fasting plasma glucose requires at least eight hours with nothing but water. An A1C requires no fasting at all and can be taken at any time of day, which is one reason it is often preferred for screening. If several tests are ordered together, such as a lipid panel alongside your sugar markers, the laboratory may still ask you to fast.
Lähteet
- Centers for Disease Control and Prevention — A1C Test for Diabetes and Prediabetes, 2024 — cdc.gov
- National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) — Recommended Tests for Identifying Prediabetes — niddk.nih.gov
- Technical University of Denmark — The length of your upper leg could help detect diabetes, July 2026 — press coverage
- Yoo D, Maggiore U, Jolliet O — Enhancing prediabetes and diabetes detection through a machine learning-enabled self-assessment approach — Journal of Clinical Epidemiology, 2026 — PubMed
- Arnardóttir E et al. — Using HbA1c measurements and the Finnish Diabetes Risk Score to identify undiagnosed individuals and those at risk of diabetes in primary care — BMC Public Health, 2023 — full reference
- Mugume IB et al. — Performance of a Finnish Diabetes Risk Score in detecting undiagnosed diabetes among Kenyans aged 18–69 years — PLOS ONE, 2023 — full reference
- Hamedi A et al. — Performance of the Finnish Diabetes Risk Score in screening for undiagnosed type 2 diabetes among adults without prior diagnosis: a Kharameh Cohort Study, Iran — Primary Care Diabetes, 2025 — full reference
Lisälukemista
- A1C-muunnostaulukko: muunna A1C-arvosi keskimääräiseksi verensokeriksi
- Diabetes: oireet, syyt, diagnoosi ja hoito
- Kattava aineenvaihduntapaneeli: tulosten tulkinta
- Verikokeen viitearvot: selitetty viitetaulukko
- Fasting before a blood test: rules, timing and what is allowed
Ymmärrä laboratoriotuloksiasi tekoälyn DiagMen avulla
A questionnaire can tell you it is time to get tested. It cannot tell you what the report means once it lands in your inbox. AI DiagMe turns markers such as fasting blood sugar, long-term sugar control (HbA1c) and insulin into plain-language context, so you arrive at your appointment with better questions rather than more anxiety. It is built to help you understand your results, not to diagnose you or replace your doctor.
Tulkitse tuloksesi muutamassa minuutissa
