"My TSH is normal and I am still gaining." It is one of the most common things an Indian adult on levothyroxine says in a consultation, and it usually arrives with an unstated theory attached: that the dose must be too low, and that the right dose would fix the weight.

The evidence points somewhere less satisfying and more useful. Hypothyroidism does add weight, but the amount is modest and most of what replacement gives back is fluid rather than fat. This piece covers how many kilograms are actually attributable to the thyroid, what the causal evidence says about which way the arrow runs between TSH and body weight, and what the safety literature says about using thyroid hormone to move the scale.

• Overt hypothyroidism adds roughly 5 to 7 kg, and most of it is fluid. Levothyroxine clears that fluid, but it does not function as a weight-loss treatment for established obesity.
• The causal arrow between TSH and body weight mostly runs from weight to TSH, not from TSH to weight. A mildly raised TSH in a heavier person is frequently a consequence of body size and can settle as weight comes down without any dose change.
• Indian laboratories commonly use manufacturer reference intervals derived from Western populations. A study of 378 Indian adults found a locally derived TSH upper limit of 7.28 µIU/mL, substantially higher than the manufacturer's range of 0.27 to 4.2 µIU/mL.
• Levothyroxine labels carry an explicit boxed warning against use for weight loss. The best available trial showed one to two kilograms of difference, achieved only with a suppressed TSH that carries cardiac and bone-density risk.
• When TSH is in range and weight is still climbing, the productive question is not whether the dose is right but what else is going on: insulin resistance, PCOS, sleep apnoea, medication effects, or energy balance itself.

The honest number is smaller than most people expect, and knowing it early prevents a lot of disappointment later.

Population-based case-control work in Denmark compared adults newly diagnosed with overt autoimmune hypothyroidism against euthyroid controls matched for age, sex and locality. At diagnosis, the hypothyroid group weighed on average 7 kg more than controls, and detailed body-composition analysis showed that the extra weight was mostly water rather than fat. Plummer's 1940 series of 200 untreated hypothyroid patients put the average gain at 4.6 kg, again judged to be entirely water retention in the form of myxoedema.

Two datasets separated by seventy years and very different methods converge on the same order of magnitude: overt hypothyroidism contributes something like 5 to 7 kg, and that contribution is largely fluid. It does not account for tens of kilograms, and a 2016 Indian review of hypothyroidism and obesity concluded exactly that, noting that severe obesity is generally not secondary to hypothyroidism.

Even within the normal range, thyroid status shifts weight by a few kilograms rather than transforming it. In the Danish DanThyr cohort, women with a TSH of 4.5 mIU/L weighed on average 5.5 kg more than women with a TSH of 0.28 mIU/L after adjustment for confounders. A real difference, and a small one.

If the gain is mostly water, the loss on treatment should be too, and that is what body-composition studies show.

Karmisholt and colleagues followed adults with overt hypothyroidism through a year of levothyroxine treatment with dual-energy X-ray absorptiometry. The key findings:

• Body weight fell by an average of 4.3 kg.
• Of that, 3.8 kg was lean mass, primarily water.
• Fat mass decreased only minimally and not significantly.
• Resting energy expenditure rose by about 11.6%, which is the metabolic normalisation the treatment is for, but it did not translate into large fat loss.

A provenance note is owed here: these figures reach the research behind this article through the American Thyroid Association's patient summary of that work rather than through the primary paper, so treat the exact decimals as indicative.

Other DXA work reports the same shape. Weight loss after levothyroxine is driven by reductions in lean mass and tissue water, not by meaningful reduction in fat mass. Clinically that matches what patients describe: facial puffiness and generalised swelling settle, sometimes with brisk early diuresis as water-binding glycosaminoglycans clear from tissues, while overall adiposity persists.

Most of the change happens in the first six to twelve months on an adequate dose, with wide variation between individuals. After that, levothyroxine has done what it does. It normalises metabolism and clears fluid-related weight. It does not function as a weight-loss treatment for established obesity, and no guideline treats it as one.

This is the part that reframes the whole conversation, and it is not intuitive.

Across large cohorts, TSH correlates positively with BMI within the reference range, free T4 correlates negatively, and free T3 tends to be higher in people with obesity, with the FT3 to FT4 ratio rising as adiposity rises. DanThyr found that people with a serum TSH above 3.6 mIU/L had roughly twice the odds of obesity compared with those at 1 to 1.99 mIU/L. Read casually, that looks like proof that a sluggish thyroid causes weight gain.

Mendelian randomisation testing says otherwise. These studies use genetic variants as instruments, which lets them ask which way the arrow points rather than merely observing that two things travel together. Wang and colleagues, publishing in the Journal of Clinical Endocrinology and Metabolism in 2021, ran a bidirectional analysis on large genome-wide association datasets and found that genetically higher TSH did not causally increase BMI or obesity risk, while genetically higher BMI did causally increase TSH and free T3. Later Mendelian randomisation work has reported the same asymmetry.

The intervention data agree. If adiposity elevates TSH, then TSH should fall when weight falls, and it does. In a retrospective study of 949 euthyroid adults with morbid obesity undergoing bariatric surgery, TSH decreased significantly at twelve months, with the largest reductions in those who started high-normal. The decline tracked independently with excess body weight lost.

A separate longitudinal study of 129 patients found TSH falling from 3.3 to 2.1 µU/mL at twelve months. In children, hyperthyrotropinaemia associated with obesity normalised after weight loss achieved through lifestyle change alone, with no thyroid treatment involved.

The proposed mechanisms are consistent with this:

• Hyperleptinaemia in obesity stimulates hypothalamic thyrotropin-releasing hormone, which raises pituitary TSH secretion.
• Adipose tissue shows altered deiodinase activity with increased conversion of T4 to T3.
• Adipocyte cytokines may subtly interfere with thyroid signalling.

None of that is thyroid failure. It is the thyroid axis responding to body size.

For an Indian reader whose TSH sits slightly above the lab's upper limit and whose weight has risen, this has a concrete implication: the weight may be contributing to the TSH reading rather than the other way round, and the reading can settle as weight comes down without any change in dose. That is a clinician's judgement to make on the individual case, not a conclusion to draw from a single report.

Before deciding a TSH is high, it is worth asking what it is being compared against, because the answer is often an imported number.

Many assay manufacturers list a reference interval of roughly 0.27 to 4.2 µIU/mL. A 2013 study assessed whether the Roche Cobas interval transfers to an Indian urban population and found that it does not.

The Roche Cobas manufacturer reference range for TSH is 0.27 to 4.2 µIU/mL. In comparison, a 2013 study of 378 Indian urban adults found a TSH reference range of 1.134 to 7.280 µIU/mL. Another study of 1,916 reference adults in Delhi reported a mean TSH of 2.2 ± 0.9 mIU/L.

Country-wide paediatric reference work across 13 states found evidence against narrowing the TSH reference range rather than for it.

The limitation is real and should be stated: comprehensive national reference intervals across different assay platforms and age strata do not exist for India. What the available data support is scepticism about a mildly raised value read against a Western manufacturer's range, not a new number to use instead.

Adequacy is defined biochemically, and the targets are age-stratified rather than universal. The ICMR Standard Treatment Workflow for hypothyroidism sets target TSH for treated patients at:

• Young adults: 1 to 2.5 mU/L
• Middle-aged adults: 1.5 to 3 mU/L
• Older patients: More permissive upper limits deliberately, to avoid overtreatment and cardiac risk

Internationally, the position is similar in shape. The 2014 American Thyroid Association guidelines keep levothyroxine monotherapy as standard of care with TSH as the primary marker of adequate replacement, targeting the reference range in non-pregnant adults and explicitly declining to recommend narrowing the upper limit to 2.5 mIU/L in general practice.

For subclinical hypothyroidism, where TSH is raised but free T4 is normal, Indian consensus recommends treating above 10 mIU/L. In the 4.5 to 10 mIU/L band, treatment is considered only where a specific factor is present:

• Pregnancy or pregnancy planning
• A large goitre
• Positive TPO antibodies
• Cardiovascular disease or dyslipidaemia
• Infertility
• Depression or refractory anaemia

Weight is not on that list, and that omission is deliberate.

Two randomised trials found no benefit from treating subclinical hypothyroidism in older adults:

• TRUST trial: 737 adults aged 65 and over, one year. No significant difference in hypothyroid symptom score, tiredness, or weight.
• IEMO 80-plus trial: Adults aged 80 and over. Same conclusion.

Guidelines also anticipate the situation this article opened with, where symptoms persist despite an in-range TSH. The instruction is to exclude alternative causes such as anaemia, sleep disorders, depression and other endocrine conditions before attributing anything to replacement adequacy, and specifically to avoid blind dose escalation that pushes TSH below the reference range.

Levothyroxine labels carry an explicit boxed warning against this exact use. The US prescribing information states that thyroid hormones should not be used for the treatment of obesity or for weight loss, that doses within the normal replacement range are ineffective for weight reduction in euthyroid patients, and that higher doses may produce serious or life-threatening toxicity, particularly in combination with sympathomimetic appetite suppressants.

The harms of over-replacement are specific rather than theoretical:

• Atrial arrhythmias and angina
• Raised heart rate and increased cardiac wall thickness
• Decreased bone mineral density
• Postmenopausal women are particularly exposed on the bone-density risk

Subclinical hyperthyroidism (a suppressed TSH with normal free T4) is associated in epidemiological cohorts with increased atrial fibrillation risk and reduced bone density.

Set that against what the weight benefit actually is. A double-blind randomised trial of 141 patients compared levothyroxine alone against a 5:1 levothyroxine plus liothyronine combination:

• Combination arm: approximately 1.7 kg weight loss on average
• Monotherapy arm: approximately 0.1 kg weight gain
• Median TSH in the combination arm: 0.07 mU/L, well below the lower limit of normal

Patient preference for combination therapy correlated with the weight change rather than with symptom scores. A smaller crossover trial found about 2.1 kg of loss over four to five months without addressing long-term cardiac or skeletal safety.

So the trade on offer is one to two kilograms in short trials, bought with a suppressed TSH and the cardiac and skeletal risk that goes with it. Both the American and European thyroid associations recommend against routine combination therapy outside formal trials, and Indian guidance does not endorse it for weight control or residual symptoms in routine practice. None of this is a dosing instruction: nobody adjusts a levothyroxine dose except the prescriber, on the basis of a blood test taken at the right interval.

There is a version of this article that would assert that levothyroxine is widely misused for weight loss in India and widely over-prescribed. The claim is plausible and it is also not evidenced.

Peer-reviewed data on the following are limited or absent in India:

• How often levothyroxine is prescribed without clear biochemical hypothyroidism
• How frequently suppressed TSH is present in Indian practice
• Rates of atrial fibrillation or fracture attributable to over-dosing

The risk estimates in the previous section come from international cohorts, not Indian data.

What Indian guidance does say is directional rather than quantified: dosing should be individualised and titrated on TSH and free T4, with lower starting doses in elderly or cardiac patients, and treatment of subclinical hypothyroidism should be selective rather than routine precisely to avoid unnecessary exposure. That guidance exists because the concern is real. The size of the problem in India is a known unknown, and it is better to say so than to fill it with a number from somewhere else.

This is where the reader who opened this article usually is: treated, biochemically adequate, still gaining. The evidence says to look outside the thyroid, and it says where to look.

Insulin resistance and PCOS

In Indian studies of women with PCOS, insulin resistance is reported in 44 to 70% using surrogate markers. One Indian study found HOMA-IR above 3 in 66.7% of cases, while another found it in 50% of lean women and 82% of overweight women with PCOS.

Obstructive sleep apnoea

Obstructive sleep apnoea is strongly associated with obesity through disrupted sleep, appetite change and impaired energy expenditure. Indian prevalence data specifically in hypothyroid populations are sparse.

Menopause transition

In women aged 40 to 55, rising central adiposity, lean-mass loss and altered energy expenditure can occur independently of thyroid status. Indian data are limited, so global patterns are being applied.

Medication effects

Certain antidepressants and antipsychotics, some antiepileptics and long-term corticosteroids are documented causes of weight gain independent of thyroid status.

Other endocrine causes

Other endocrine causes include Cushing's syndrome, hypogonadism, and rare hypothalamic or genetic obesity syndromes. Each requires targeted evaluation based on clinical signs rather than routine screening.

Energy balance itself

Energy intake is systematically under-estimated and activity over-estimated, particularly while dieting. This can continue even after metabolism has normalised on levothyroxine.

None of these factors is self-diagnosable, and several require referral or targeted evaluation rather than routine screening. PCOS also requires additional context, since it has four distinct presentations, and the significance of insulin resistance can differ depending on the presentation.

These factors shift the conversation from "is my dose right", which the TSH has already answered, to "what else is going on", which it has not.

Long-term weight management in a euthyroid person with treated hypothyroidism looks like weight management in anyone else. It is worked out against diet, activity, sleep and comorbidities rather than against further increments of thyroid hormone.

The person best placed to run that assessment is a clinician who can read the thyroid results and the rest of the picture together. If you are choosing where to have that conversation, what endocrinologist-led means sets out what to check before you accept the label at face value.

TSH in range, weight still climbing? A metabolic assessment looks at the full picture. A Voy clinician reads the thyroid results and everything else together. Book an assessment.

This article is for general information and education only and is not medical advice, diagnosis, or treatment. GLP-1 and other medications referenced are prescription-only and are appropriate only for certain people under the supervision of a qualified clinician. Do not start, stop, or change any medication based on this article. Please consult a registered medical practitioner about your individual circumstances. Information reflects what was available at the time of review and may change.