Nausea, reflux, constipation and loose stools are what most people on GLP-1 treatment notice first, and on an Indian plate they tend to announce themselves at three predictable points: the size of the serving, the fat carried in the tempering, and the fibre load of a dal-heavy meal.

Most of it can be managed by changing what reaches your plate rather than what goes into the pot, and most of it is worst in the early weeks or after an increase. Common is not the same as something to be endured. What follows is what the evidence supports, what is only convention, and where an ordinary symptom becomes a reason to call a clinician.

Key Takeaways:

• Nausea typically lasts 8 days per episode. Constipation lasts a median 47 days, which is much longer and worth addressing rather than waiting out.
• The first move is portion reduction, not a menu change. The stomach effect on satiety means last month's normal serving is this month's trigger.
• The fat in an Indian meal sits mostly in the tadka and finishing ghee, not the base dish. Setting aside a portion before the final tempering is easier than redesigning the whole meal.
• Fibre pulls in two directions: gradual additions of soluble fibre (dal, oats, isabgol) help constipation; adding large amounts of high-fibre foods when diarrhoea is the symptom can worsen it.
• Any symptom that prevents keeping fluids down, or that arrives with severe abdominal pain radiating to the back, needs same-day assessment rather than dietary adjustment.

GLP-1 medicines slow the rate at which the stomach empties and increase fullness, so a meal sits longer and a portion that was unremarkable last month can produce early fullness, nausea, belching or reflux. Nausea is not explained by slowed emptying alone, which is why a smaller meal helps some people and not others, and the intestinal mechanisms behind constipation and diarrhoea remain poorly defined.

The stomach effect does not necessarily persist. In a 2021 randomised study of 72 adults, semaglutide showed no detectable emptying delay at week 20 when measured indirectly through paracetamol absorption, although that test has known limitations. Symptoms can nevertheless carry on after the measurable stomach effect has waned, so the absence of a delay on a test is not a promise that the nausea has finished.

One thing worth settling before you start rearranging meals is whether an existing gut condition changes the calculation, because some digestive diagnoses sit among the reasons GLP-1 treatment may not suit a particular person in the first place. That is a prescriber's assessment, not a self-assessment.

The opening month is a recognisable trouble window rather than a vague warning. In a 2017 analysis of 2,134 trial participants, first gastrointestinal events clustered in the first five days, and liraglutide-related nausea, vomiting and diarrhoea ran highest across weeks zero to four.

How long an individual episode lasts is a separate question from how long the tendency lasts. A 2022 pooled analysis of 3,379 participants in a semaglutide programme found these median episode durations:

Median episode duration by symptom (2022 pooled semaglutide data, midpoints, not individual forecasts):

• Nausea: 8 days
• Diarrhoea: 3 days
• Vomiting: 2 days
• Constipation: 47 days, the one that outstays every other symptom by a wide margin

The pattern is still useful even as a midpoint: the symptoms people dread most are usually the shortest-lived, while constipation is the one that requires active attention rather than patience. That contrast should shape how you handle fibre, which is dealt with further down.

Symptoms commonly return after a dose increase, which is one reason the pace of any increase is a prescriber's decision and not a calendar entry.

The first adjustments are about volume and speed, not ingredients, and they are worth making before anyone reorganises the kitchen:

• Halve the serving: Take half your usual rice, roti, dal or sabzi, and treat the second helping as optional rather than default.
• Eat slowly and stop early: Stop at the first point of comfortable fullness rather than at the end of the plate, since the fullness signal arrives late when the stomach is emptying slowly.
• Spread the day out: Smaller, more frequent meals are generally better tolerated than two or three large ones.

State the evidence grade honestly here, because it matters. Smaller portions, slower eating and increased meal frequency are expert-consensus measures carried across gastroenterology and dietetic guidance. They have not been tested in GLP-1-specific food trials, and no trial has established an optimal meal size, meal frequency or eating rate for someone on this treatment. They are sensible and low-risk, which is a different claim from demonstrated.

Fat is the one dietary lever with a demonstrated mechanism behind it. High fat content measurably slows gastric emptying, which is precisely the effect the medicine is already producing, so a fried or richly tempered meal stacks one delay on another. Its effect on reflux specifically is less consistent in the literature, so treat fat reduction as a nausea and fullness measure first.

The difficulty is that fat in Indian cooking is rarely visible on the plate. It arrives in the tempering, the shallow-frying and the finishing ghee rather than as a slab of anything. The Indian Nutrient Databank estimates these figures per 100 g of prepared recipe:

Fat per 100 g prepared (used as a ranking between preparations, not as precise per-dish values):

• Idli: 0.33 g, the lowest fat per gram in common Indian preparations
• Chapati: 3.56 g, modest without significant added ghee
• Plain dosa: 8.43 g, rises sharply with oil. A well-wiped pan makes a material difference
• Aloo paratha: 10.22 g, the highest in this comparison

Two caveats sit on those figures and both are load-bearing. These are recipe calculations built mainly on ingredient composition data rather than laboratory assays of every cooked dish, and household oil practice can shift them substantially in either direction. A dosa made in a lightly wiped pan and a dosa swimming in oil are not the same food, and no database can tell you which one your kitchen produces.

Used as a ranking rather than as precise values, the figures point the same way as the mechanism. Only the four preparations above carry measured values from that source. Everything named below is placed by cooking method rather than by a databank figure.

During a bad stretch, the steamed and lightly tempered end of the repertoire is the easier end: idli, plain rice, phulka without ghee, lightly tempered dal, and boiled or grilled preparations. Poori, paratha, pakora and thick, rich gravies are the ones most likely to sit heavily. This is a temporary adjustment for a symptomatic period, not a verdict on any of these foods.

Fibre advice fails on GLP-1 treatment when it is applied as a single instruction, because the two most common bowel symptoms need opposite handling.

For constipation: Add fibre gradually, with fluid, from oats, fruit, vegetables, pulses, or isabgol. Across 16 trials in non-GLP-1 chronic constipation covering 1,251 adults, published in 2022, fibre supplementation improved response rates but also increased flatulence. That trade-off is why you add fibre slowly rather than in one enthusiastic week. Someone already bloated on treatment may find a sudden fibre push makes the day worse before it makes it better.

During active diarrhoea: The direction reverses. Coarse bran, large pulse portions and sugar alcohols are reduced temporarily, then restored gradually once stools settle. That sequence is consensus advice rather than trial-derived, and it has not been tested in people on GLP-1 treatment.

One claim you will see confidently asserted and which this piece will not make: that Indians eat too little fibre and should therefore add more by default. No current nationally representative Indian dataset cleanly establishes adult fibre intake, so the honest position is that your fibre change should follow your symptom rather than a national average nobody can currently measure.

Three household beliefs come up constantly in this conversation, and they deserve separately calibrated answers.

Chilli is personal, not universal: Capsaicin can provoke burning, and that is real, but the evidence is mixed and none of it is GLP-1-specific. A 2020 crossover study in 15 patients with reflux found more abdominal burning after red chilli but only a trend toward increased heartburn. In a study that small, that is not a finding to build a rule on. Treat your own tolerance as the test rather than removing chilli reflexively.

Reflux responds to timing more reliably than to spice: Finishing dinner two to three hours before lying down is the standard measure, and it is worth knowing that major GERD guidance rates this as a conditional, low-certainty recommendation. It costs nothing, which is a fair reason to try it.

Curd is reasonable, not therapeutic: Plain curd is a perfectly sensible food if you tolerate it, and it is a soft, low-fat option when appetite is poor. Ordinary homemade curd has not been shown to treat GLP-1 gastrointestinal symptoms, and probiotic benefits in the wider literature are strain-specific, which household dahi is not. If dairy worsens your diarrhoea, do not push through it on the strength of a general belief that curd settles the stomach.

Most Indian readers are not cooking for one, and advice written for a single-portion kitchen quietly assumes control that does not exist. The practical move is not to change the family menu but to change where your portion is taken from.

Ask for your share to be set aside before the final tadka, the finishing ghee or the cream goes in. The dish that reaches the table is unchanged for everyone else, and yours has lost the fat layer most likely to sit heavily.

The same logic covers the rest of it: your rice is served first and smaller, your dal is taken from the pot before it is enriched, and nobody is required to cook two dinners. Where a household eats late by habit, moving your own meal earlier is usually easier to negotiate than moving everyone's.

The line between manageable and clinical is not about how unpleasant a symptom feels. Contact your prescriber if vomiting or diarrhoea persists, if fluids will not stay down, if urine becomes scant or dark, or if constipation or reflux keeps worsening despite the measures above.

Seek urgent care for severe, persistent abdominal pain, particularly pain that reaches through to the back, with or without vomiting. Regulatory product information identifies this as a possible pancreatitis warning.

Nothing on this page is a reason to change, pause or stop treatment on your own. The reasonable thing to do with a symptom that is not settling is to bring it to the clinician who prescribed the medicine, with a note of when it started, what you have tried and what you can actually eat. That is the information a clinical review is built on.

Side effects persisting beyond what ordinary dietary adjustments can reach? Speak to a Voy clinician. A review may adjust timing, dose or the plan around them.

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.