Losing weight without trying

What counts as weight loss worth investigating

The conventional threshold, and the one used here, is a loss of more than about 5% of body weight over six to twelve months that was not intended — around 3 kg for a 60 kg adult. Below that, and over a longer period, weight fluctuates for ordinary reasons. Above it, something is usually driving it.

Two practical difficulties come up constantly. The first is that most people do not weigh themselves, so the history is "my clothes have become loose" or "people are telling me I look thin" rather than a number. That is still usable evidence, and a belt notch or a blouse that no longer fits is worth mentioning. The second is that weight loss is often reported as weakness, tiredness or loss of appetite instead — which is why this page and the weakness guide are read together.

Why tuberculosis comes first here

In a low-incidence country, unexplained weight loss in an adult prompts a search for malignancy first. In Cooch Behar that ordering is wrong, and the reason is the base rate. India's national TB prevalence survey found microbiologically confirmed pulmonary tuberculosis in 316 per lakh of the population aged 15 and over,^[1] with a national notification rate of 178.8 per lakh in 2023.^[2] Tuberculosis is not a rare disease being hunted; it is the likeliest single explanation in front of you.

The part that gets missed is that TB does not have to announce itself with a cough. India's National TB Elimination Programme defines a presumptive patient as one with "cough for 2 weeks or more, fever for 2 weeks or more, significant weight loss, haemoptysis, any abnormality in chest radiograph" — weight loss is listed in its own right.^[3] And the evidence for why that matters is unambiguous: cough of two weeks or more, used as a screening test on its own, has a pooled sensitivity of just 0.42 (95% CI 0.36–0.48) across 40 studies and 6,737 patients with TB.^[4] Fewer than half of people with tuberculosis report the symptom everyone screens on.

So the person losing weight with evening fevers and no cough at all still gets sputum sent and a chest X-ray. If sputum cannot be produced, that is a reason to investigate differently rather than a reason to stop.

Two further points on testing. Sputum smear microscopy misses a great deal: on the same specimens in the national survey, CBNAAT found 825 cases where smear found 445.^[1] And tuberculosis outside the lungs is common here — abdominal tuberculosis accounts for roughly 13% of extrapulmonary TB in India against 6% in the United States, and presents with abdominal pain, weight loss and altered bowel habit rather than anything respiratory.^[5] A normal chest X-ray does not end the question.

Diabetes, which announces itself this way

New or badly controlled type 2 diabetes causes weight loss, and in a meaningful number of people that is the first thing noticed — before the thirst, before the frequent urination, and certainly before any complication. The mechanism is simple: glucose is being lost in the urine, taking calories with it, and without enough insulin action the body breaks down fat and muscle instead.

This is worth emphasising because the popular association runs the other way. Diabetes is thought of as a disease of being overweight, so weight coming off is read as an improvement rather than a warning. Someone with known diabetes who is losing weight without trying should have their control checked promptly, not congratulated. An HbA1c and a fasting glucose are cheap, quick, and among the first tests done. The diabetes guide covers diagnosis and targets.

The thyroid

An overactive thyroid produces weight loss with a good or increased appetite, together with palpitations, tremor, heat intolerance, anxiety and loose stools. It is frequently mistaken for anxiety or simply for "nerves", particularly in younger women. It is one blood test. No reliable population prevalence figure for thyrotoxicosis in India exists that could be verified for this page, so no number is offered — but the test is inexpensive and it is done routinely in this situation because the diagnosis is easy to miss and straightforward to treat. Interpretation is covered on the thyroid page.

Working through the rest

Alcohol, said plainly

Alcohol contributes to weight loss through liver disease, pancreatic disease and simple displacement of food. In West Bengal, alcohol-use disorder affects about 3.04% of adults and the treatment gap is 89.2% — nine in ten people affected receive no treatment, with a mean delay of five years before any is sought. The female-to-male ratio in West Bengal is 1:5, against 1:18 nationally.^[8] A Kolkata series of 395 patients with alcohol-related liver disease found 79.7% already had cirrhosis or acute-on-chronic liver failure when they first presented.^[9]

The question is asked here without judgement and the answer is not written down anywhere that anyone else sees. It changes what is tested and what can be offered, which is the only reason it is asked.

Cancer, in proportion

Almost everyone losing weight is afraid of cancer, and that fear deserves a straight answer rather than reassurance or evasion. Malignancy is a real cause and it is looked for. But two facts should be set against the fear.

First, in this region tuberculosis is a great deal commoner than cancer as a cause of weight loss, which is why it is excluded first. Second, the cancers people most fear are not the ones with the highest incidence here. In the Kolkata population-based registry, stomach cancer incidence is 4.2 per 100,000 in men and 2.4 in women. In Aizawl, Mizoram, the same figures are 44.2 and 21.7 — more than ten times higher.^[10] Bengal is a low-incidence region for stomach cancer.

Where North Bengal does show a distinct pattern is gallbladder cancer. Among 2,058 cancer cases at North Bengal Medical College, Siliguri, between 2010 and 2012, gallbladder cancer accounted for 6.5%, rising from 3.8% to 7.8% across those three years, with 93.3% in women.^[11] That is a hospital case series rather than a population registry — it describes what one referral centre saw, not the incidence in the district — but weight loss with right upper abdominal pain in a woman here is a combination that gets imaged rather than observed.

It should also be said honestly that Kolkata is the only population-based cancer registry in West Bengal. There is no registry for North Bengal, so no true incidence figure for this district exists for any cancer. Anyone quoting one is estimating.

How far to investigate

There is a real tension here. Weight loss without any localising symptom can lead to open-ended testing that is expensive, anxiety-provoking and often unrewarding — and where cost is a genuine constraint, as it is for most people attending this clinic, a scattergun approach means the useful tests get delayed by the useless ones. The approach taken is to do a defined first round, then let the results and any new symptom direct the second, rather than ordering everything at once.

The first round is: full blood count, ESR or CRP, HbA1c and fasting glucose, thyroid function, creatinine and eGFR, liver function with albumin, urine analysis, chest X-ray, sputum for molecular TB testing where there is any respiratory or febrile element, and an HIV test. In West Bengal, where 71.4% of women and 38.9% of men aged 15–49 are anaemic,^[12] the blood count is doing real work rather than making up numbers.

If that round is normal and the weight is still falling, that is a reason to look further rather than to stop — but the next step is chosen from the history, not from a list. Where cost is a constraint, say so at the consultation: the tests can be prioritised so the ones most likely to change the treatment are done first. Follow-up within 10 days is not charged, which exists precisely so that coming back to have reports read does not cost a second consultation.