Low-Calorie Snacks: What the Evidence Says, and What Changes on a GLP-1

Snacking does not make people gain or lose weight on its own: in randomized trials, eating more often and eating less often produce about the same weight change.^[1][2] What a snack is made of matters more than whether you eat one. Watery, low‑energy‑density foods and protein-rich snacks keep people full for longer per calorie than crackers or chocolate,^[6][9] and nuts, despite their calories, have not caused weight gain in trials.^[10] On Ozempic, Wegovy, Mounjaro or Zepbound the question changes. The drug already cuts intake by about a quarter or more,^[14][15] so the useful snack is rarely the one with the fewest calories. It is the one with the most protein, and later the most fiber, for the calories it costs.

The evidence at a glance

Does snacking help or hurt weight loss?

The old idea that small, frequent meals “stoke the metabolism” has been tested and has not held up. A 2023 meta-analysis of randomized trials compared eating three or fewer times a day with four or more. The difference in weight change was 0.62 kg, with a confidence interval running from 2.76 kg less to 1.52 kg more, and 15 of the 16 trials were at high risk of bias. The authors could not recommend either pattern.^[1]

A 2024 meta-analysis in JAMA Network Open found a slight edge for fewer meals, about 1.85 kg over at least 12 weeks, and called the effect small and of uncertain clinical importance, again with most trials at high risk of bias.^[2] The cleanest single test held calories fixed: 16 adults with obesity on the same deficit for eight weeks, one group eating three meals and the other three meals plus three snacks. Weight, body fat, appetite and gut hormones did not differ.^[3]

The scientific committee advising on the 2025–2030 Dietary Guidelines for Americans reviewed snacking and energy intake in adults and found that no conclusion could be drawn, because the five available trials were too narrow to apply to the general public.^[4] What the observational literature does show is narrower: in adults followed over time, more “discretionary” snacking, meaning chips, sweets and similar foods, went with rising weight, while snacking in general showed no consistent link.^[5] How often you eat looks roughly neutral for weight. What you eat does not.

What makes a snack filling: energy density

Energy density is calories per gram. Foods heavy in water and fiber, such as fruit, vegetables and broth-based soup, sit low; dry, fatty foods such as crackers, chips and chocolate sit high. Barbara Rolls’s group at Penn State has spent decades showing that people tend to eat a fairly steady weight of food, so lowering energy density lowers calories without a fight with hunger.

Two of their year-long trials tested this directly. In one, 200 adults followed the same calorie-reduced plan and were randomized to add two servings a day of low-energy-density soup or the same calories as high-energy-density snack foods such as crackers and pretzels. The soup group lost 7.2 kg; the snack group lost 4.8 kg.^[6] In the other, 97 women with obesity were told to cut fat, with or without adding water-rich fruit and vegetables, and had no calorie targets. The fruit-and-vegetable group lost 7.9 kg against 6.4 kg, ate a greater weight of food, and reported less hunger.^[7] Both trials come from one research group, and neither was about GLP‑1 drugs.

Protein snacks and how long they hold you

Protein is the macronutrient most consistently linked to fullness, and the snack trials support it within limits. In one, 15 young women ate a 160-calorie afternoon yogurt with 5, 14 or 24 g of protein, or no snack. Without a snack they asked for dinner after about two hours; with the highest-protein yogurt, almost three hours. Every snack reduced hunger, and dinner was smaller after a snack, but not by enough to cancel the snack’s 160 calories.^[8]

A follow-up compared the same calories as a high-protein yogurt, crackers or chocolate in 20 women. The yogurt cut afternoon hunger more than the chocolate, pushed the next meal back by about 30 minutes, and was followed by roughly 100 fewer calories at dinner than the chocolate.^[9] These are single-afternoon tests in lean young women. They show what holds people longer, not that protein snacks produce weight loss.

Nuts: dense in calories, neutral for weight

Nuts are the classic case where calorie counting and trial results disagree. An ounce of almonds is about 164 calories by the USDA figures,^[17] yet a 2021 meta-analysis of 86 randomized trials found no effect of eating nuts on body weight (a difference of 0.09 kg), graded high certainty, and cohorts of more than half a million people linked nut eating with slightly less overweight and obesity.^[10] Several of its authors disclose funding from nut trade bodies. An independent meta-analysis of 55 trials reached the same result whether or not people were told to swap the nuts for other food.^[11]

Two findings explain part of this. In a four-week trial in 137 adults given 43 g of almonds a day, people ate less of everything else without being told to, and weight did not change.^[12] And not all of the label calories are absorbed: a controlled feeding study measured almonds at 4.6 calories per gram, or 129 per 28 g serving, against the 168 to 170 predicted by the standard formula.^[13] Nuts are a reasonable snack on weight grounds. They are a poor protein source per calorie, which matters on a GLP‑1.

What changes on a GLP-1

GLP‑1 drugs act directly on snacking. In a 12-week crossover trial, semaglutide 1.0 mg lowered food intake at lunch, at the evening meal and at snacks, adding up to 24% less food across the day, and lowered the relative preference for high-fat foods.^[14] At the 2.4 mg Wegovy dose, intake at a buffet lunch was 35% lower than on placebo after 20 weeks.^[15] The 2025 joint advisory from the American College of Lifestyle Medicine, the American Society for Nutrition, the Obesity Medicine Association and The Obesity Society puts the observed calorie reduction at 16% to 39%.^[16]

That shifts the risk. The same advisory warns that vitamin and mineral intake becomes hard to cover below about 1,200 calories a day for women and 1,800 for men, and that weight lost on these drugs includes lean tissue: in the STEP 1 trial of semaglutide, 5.3 kg of an average 13.6 kg lost (38%) was lean mass.^[16] It notes protein targets of 1.2 to 1.6 g per kilogram a day have been proposed during active weight loss, against the general 0.8 g/kg allowance.^[16] How those targets are set and when they do not apply, including kidney disease, is covered in protein targets and the lab panel; what runs short in practice is in eating too little on a GLP-1.

What follows is reasoning, not a trial result. Someone whose appetite is already suppressed does not need a snack that is low in calories. They need one that fills a gap in protein, and once the stomach has settled, fiber, without taking much room. The advisory’s own eating-habit list points the same way, encouraging regular, small meals at consistent times and discouraging long stretches without food, while listing sweets and savory snack foods among things to minimize.^[16]

Snacks compared by protein per calorie

The table below uses USDA FoodData Central figures for plain, unbranded foods, multiplied out to common portions.^[17] The last column is the one that matters on a GLP‑1: grams of protein for every 100 calories. Brand products, flavored yogurts and seasoned jerky will differ, so check the label.

The plain dairy and lean-meat snacks deliver four to six times the protein of nuts for the same calories. Raspberries, apples and popcorn are where the fiber is, but they carry almost no protein. And the snacks most often sold as “low calorie” are not always the lightest: jerky is energy-dense, and the nuts are the most calorie-heavy items on the list. Pairing helps. Yogurt with raspberries, or cottage cheese with an apple, covers protein and fiber in about 160 to 180 calories. Plant options are compared in plant-based protein on a GLP-1.

What “low calorie” on a package means

On a US food label, “low calorie” is a defined claim. For most foods it means no more than 40 calories per reference amount customarily consumed, a standard serving size set by regulation.^[18] Almost nothing in the table above qualifies, and most foods that do (cucumber, celery, plain broth) carry little protein. A package that says “100-calorie pack” is making a portion statement, not a nutrition one. On a GLP‑1, the more useful figure on the label is the protein line.

Putting it to use on a GLP-1

• Snack if it helps you reach protein, not out of habit. If meals already cover your protein target, a snack is optional. The protein calculator gives a daily figure to aim at.
• Lead with protein. Tuna, deli turkey, plain Greek yogurt and cottage cheese give 17 to 22 g per 100 calories. Eggs, string cheese and edamame give about half that.
• Add fiber once nausea settles. Berries, an apple, edamame or air-popped popcorn. The fiber calculator estimates a daily target.
• Avoid long gaps. Going so long without food that you become overly hungry is on the advisory’s list of habits to minimize.^[16]
• Treat nuts as a fat-and-fiber snack, not a protein one. They do not cause weight gain in trials, but at about 6 g of protein per ounce, they take up room a GLP‑1 appetite may not have.
• Skip the chips and sweets for the reason the trials give. Energy-dense snack foods were the arm that lost the least weight in Rolls’s year-long trial.

For whole-day eating rather than snacks, see what to eat on Ozempic. If between-meal cravings are the problem rather than low intake, food noise covers what the drugs do to them.