Weigh once a week and this works out whether the change matters. Losing weight on purpose is judged one way. Losing it without meaning to — especially after 65 — is treated as a warning sign, not a win.
Weigh once a week. Answer a few questions about the person. This page then tells you whether the change is worth taking to a doctor.
Start with an age and today's weight. You do not have to fill in everything — age, usual weight and two weigh-ins give a useful answer; the rest sharpens it.
This page cannot diagnose anything and does not try to. What it does is work out whether an appointment is worth having, and make sure that appointment is a good one.
If section 08 says clinical review, book it — do not wait for another weigh-in to be sure. Before you go, press Build clinician summary in section 10. It writes out the weights, the percentages, the timeline and every factor you ticked, in one block you can print or paste into a message. Serial weights over months are the finding; a single reading is not.
If it says keep weighing, that is a real answer too. Carry on with the weekly weigh-in and check back in a few weeks.
That is the common case, and everything works the same way — use their age, their weights and their answers throughout, not yours.
Two things are worth checking rather than assuming. Ask what they used to weigh, since usual body weight is the anchor for every percentage on the page. And check clothing fit and belt notches yourself: weight loss is often played down, especially when it was welcome, and there is a box for exactly that in section 06.
Your entries are saved in this browser on this device. When the page computes an assessment, the figures you have entered — profile, weigh-ins, screen answers and settings — are sent over HTTPS to the MedConsults server, processed transiently for that single computation, and never stored or logged there. Nothing else is sent anywhere. Your saved data is not backed up: if you clear your browser data or move to another machine, it is gone.
Use Back up (JSON) in section 10 to save a copy first, and Restore backup to bring it back.
Start here. Your answers change every rule the tool uses after this point.
Half a kilo lost in a week is a good week if you are trying to lose weight. The same half kilo is a reason to ring the doctor if nobody planned it.
After about 65, weight loss takes muscle with it whether or not it was intended — and muscle is what keeps people steady on their feet, out of hospital, and able to recover from illness. So the same drop on the scale is read far more cautiously in an older adult.
Under 65 and losing weight on purpose, the tool uses the NIH weight-management brackets. From 65 up it switches to the geriatric standard, where weight counts as a vital sign.
One nuance worth holding onto: weight naturally peaks around 55 in men and around 65 in women, and drifts gently down afterwards. A slow decline in later life is not automatically a warning. Rapid decline is, and that is the distinction this page is built to make — which is also why unintended weight loss is often one of the first visible signs of dementia, years before memory problems are obvious.
Fill this in once. Come back and change it only if something big changes.
The most important box here is usual body weight: what the person steadily weighed before any of this started. Every percentage in this tool is measured against that number, so a rough but honest figure beats leaving it blank. It is not the same as the first weight you record today.
Weigh the same way every time, or the numbers will not compare.
Same day of the week, first thing in the morning, after using the toilet, before eating or drinking, no shoes and no heavy clothes. Only the weight and the date are needed — everything else on this list is a bonus.
A healthy adult's weight wanders by up to about half a kilo (roughly a pound) from day to day without anything changing. A jump of 1–2 kg overnight is water, not fat or muscle — nobody can eat or burn that much that fast.
That is why one weigh-in tells you almost nothing and six tell you a great deal. If you enter a date you have already used, the new figure replaces the old one.
Two or three days a week, working each major muscle group — legs, hips, back, abdomen, chest, shoulders, arms. Weights, resistance bands, or your own body weight all count, as do heavy gardening and climbing. Two to four sets, and ten to fifteen repetitions is the range recommended for middle-aged and older people starting out; fewer, heavier repetitions build more raw strength but are not the place to begin.
Leave at least 48 hours before working the same muscle group again — that gap is when the adaptation actually happens. Start at a very light or light intensity if the person has been sedentary.
Two things to check with a doctor first rather than assume. Resistance training is not appropriate without medical advice in poor left-ventricular function, unstable angina, proliferative diabetic retinopathy, cardiac autonomic neuropathy, a history of exercise-induced arrhythmia, or vertigo and vestibular problems. And anyone with diabetes on insulin or a secretagogue should check blood glucose before and after: below 100 mg/dL beforehand, take about 15 g of carbohydrate first.
The aerobic side is separate and sits on top: 150 minutes a week of moderate activity, the same target at 65 and over as under.
Losing 5% of your weight matters a lot in one month and much less over a year. So the tool checks each length of time separately.
Each box below is one span of time with its own limit. A box lights up when the loss over that span reaches the limit doctors treat as significant.
They are the malnutrition criteria published by ASPEN and the Academy of Nutrition and Dietetics: 5% in a month, 7.5% in three months, 10% in six months, 20% in a year. For older adults, 5% over six to twelve months is the widely accepted marker on its own.
Standard family-medicine teaching uses the same three anchors — 5% in one month, 7.5% in three, 10% in six, measured from usual weight — and treats anything beyond those figures over the same interval as severe rather than merely significant. Either way, crossing a line here means the cause needs finding, not that a diagnosis has been made.
Percentages are measured against usual body weight where you have given one, otherwise against the nearest earlier weigh-in.
Those four windows are all about speed. There is a second question that speed cannot answer: how far below normal the person has actually ended up. Someone who drifted down slowly for two years and someone who dropped sharply in two months can arrive at exactly the same place, and it is the place — not the journey — that decides how urgent the nutrition response is.
So section 02 also reports current weight as a percentage of usual body weight, staged the way clinicians stage it: 85–90% is mild depletion, 75–84% is moderate, and under 75% is severe. Below about 70% — that is, once roughly 30% of body weight has gone — feeding has to be rebuilt gradually and under supervision, because a sudden return to full intake can drop blood phosphate far enough to affect the heart.
Scales wobble. Before treating a drop as real, it is worth asking whether it is big enough to be more than the wobble.
How much you drank, whether you have been to the toilet, what you are wearing and the scale itself can shift the reading by about a kilogram in a body that has not changed at all. In a small adult that is roughly the size of the 5% threshold — so the wobble alone can look like a warning.
Instead of joining up the last two readings, it fits a line through all of them, so one odd morning does not swing the answer. It then reports how wide the uncertainty around that line is.
It also runs a drift monitor, which adds up small departures from the expected path. A slow slide can stay under every single threshold for months while still being real; adding the small departures together catches it sooner than any one cut-off can.
Go through this list whenever weight is falling. Tick anything that sounds familiar — you are not diagnosing anything, just collecting what a doctor would want to know.
Medicines, low mood, sore teeth and trouble getting to the shops explain a large share of unplanned weight loss, and every one of them can be sorted out. They are worth ruling in or out before anyone worries about anything worse.
In older adults, roughly a third of unexplained weight loss traces back to cancer, about a third to other disease, and the rest is never fully explained. That last group is larger than most people expect.
The point of a list like this is not to sort those groups out — that needs a clinician and often tests. It is to make sure the cheap, fixable causes get named out loud before anything else.
Under the same falling line on a chart there are three quite different processes, and they do not have the same answer.
Wasting is loss caused simply by not taking in enough calories. More food genuinely fixes it, and most of the items on the checklist in this section cause it.
Cachexia is loss of fat-free mass — muscle, bone, organ tissue — driven by an inflammatory, catabolic illness such as cancer or heart failure. Food alone does not fix it, because intake is not the problem. Trying to out-eat it is a common and understandable mistake.
Sarcopenia is the age-related loss of skeletal muscle, driven by falling activity, changing anabolic hormones and low-grade chronic inflammation. It responds to resistance exercise combined with protein — and notably, protein alone or exercise alone does not hold muscle nearly as well as the two together.
Telling these apart needs a clinician and usually some tests. But knowing that they exist is the reason this page never answers a falling weight with "eat less of something".
Some of the earliest signs are visible rather than measurable, and they are easy to notice at home and easy to forget at an appointment. Worth mentioning if you have seen any of them: hollowing at the temples; loss of padding around the eyes or at the back of the upper arm; rings, watches, dentures or shoes becoming loose; difficulty rising from a low chair or climbing stairs; a weaker grip on jars and taps; cuts and grazes healing slowly; easy bruising; and changes in skin, hair or the lining of the mouth, which tend to show first because those tissues renew fastest.
One more that is worth checking rather than asking about: belt notches and clothing fit. People underestimate their own weight and overestimate their own height, and welcomed weight loss in particular tends to be reported as smaller than it was.
If a nurse or doctor has given you an MNA®-SF score, enter it here. If not, you can skip this — everything else still works.
You can also fill in the questionnaire yourself, free, at mna-elderly.com, and type the score in below.
The Mini Nutritional Assessment Short Form is a short, well-tested questionnaire for older adults. It asks about food intake, recent weight loss, mobility, recent illness or stress, memory and mood, and BMI — using a calf measurement where BMI cannot be worked out. It scores out of 14.
It is a published, licensed instrument, so this tool records and tracks a score you obtain from the real form rather than printing a copy of it. The DETERMINE checklist is a reasonable alternative for people living at home.
This pulls every answer so far into one number: the chance that something is going on which is worth a doctor's time. Then it tells you whether that chance is high enough to be worth acting on.
Not every worry counts the same. New gut symptoms carry far more weight than eating alone, so the tool weighs each one rather than counting ticks. The table below shows exactly what each answer contributed, so nothing is hidden.
A probability on its own does not tell you what to do. Acting has a cost — an appointment, worry, sometimes tests — and so does not acting. The point where those two balance out is the threshold, and it depends on how much worse it is to miss something than to check unnecessarily.
For unplanned weight loss in an older adult, missing something is much worse, so the threshold is low: a fairly small chance is already enough to justify a conversation. The ratio behind it can be changed in section 11.
The review reads your trend against the standard you set in section 01 and writes the adjustment for the coming week.
Every weigh-in you have entered. Use Build clinician summary before an appointment — it writes out everything a doctor needs in one block you can print or paste into a message.
Most people can skip this section entirely. Nothing here needs changing for the tool to work.
These are the assumptions the calculations run on. None of them is measured from the person being tracked — they are starting estimates, put on the table so anyone who disagrees can change them and see what happens. Leave a box blank to keep the default shown in grey.
The published limits in sections 04 and 06 — the 5%, 7.5%, 10% and 20% thresholds, and the MNA-SF score bands — do not come from anything here and do not move when you change these boxes.
What does move is the probability in section 07 and the width of the uncertainty in section 05. If you are unsure, the honest experiment is to set a value to something you consider extreme and check whether the recommendation actually changes. Often it will not, which tells you the conclusion is robust.