Food cravings: why they strike and what actually helps

Half past three in the afternoon. Your energy dips, your head starts negotiating with the vending machine, and ten minutes later the wrapper is in the bin and the guilt arrives on schedule. If you have been telling yourself this is a willpower problem, here is the better frame: it is not. Cravings have concrete causes - short nights, packed schedules, meals without protein - and the same causes respond well to treatment. This is what the research actually shows, and what helps on an ordinary working day rather than an idealised one.

On days when lunch shrinks to a coffee, I mix a ProShape All-in-1 Wild Berries instead - thirty seconds, and the afternoon starts with protein on board rather than an empty stomach.

What is actually behind the wave

The strongest craving amplifier is probably your alarm clock. In a laboratory study at the University of Chicago, twelve healthy young men slept two nights restricted to four hours and two nights extended to ten, fed identically and otherwise treated the same. After the short nights, their leptin fell by 18 percent, their ghrelin rose by 28, and they reported 24 percent more hunger - with the appetite for calorie-dense, carbohydrate-heavy food climbing 33 to 45 percent [1]. One experiment, twelve men, no universal law. But the direction holds up at scale: a meta-analysis pooling 17 laboratory studies of partial sleep restriction calculated that participants ate around 385 extra calories a day afterwards, without burning more [2]. Someone sleeping four hours instead of eight is not simply undisciplined. They are running a hormonally damped brake with a stronger accelerator.

The second amplifier is stress. In a laboratory study, women whose cortisol spiked strongly after a stressful task ate noticeably more calorie-dense sweets than women whose cortisol barely moved - on calm comparison days, the two groups did not differ [3]. Why stress reaches for chocolate and biscuits is straightforward: sweet, palatable food dials the tension down. A body of research describes this as self-medication - a stressful stretch, a sweet reward, and the internal pressure briefly drops [4]. The body is doing nothing crazy. It is looking for the fastest descent from the stress level. The bill usually arrives a day later.

The third factor sits in your kitchen cupboard: bans. Research on eating control distinguishes rigid and flexible restraint. In a classic study, women received a calorie preload before an ice-cream taste test. Those who applied their food rules rigidly ate as much or more ice-cream afterwards; those with flexible rules ate less following the preload [5]. Dieters know this pattern as the "what-the-hell effect": one slice of cake breaks the diet, so the rest might as well go too. All-or-nothing rules turn every slip into a freefall. Flexibility turns it into a stop.

Protein is your strongest lever

Satiety is not a lottery, it has a ranking. Among the macronutrients, protein is the most reliable filler - more so than carbohydrate or fat, within a meal and across whole days [6][7]. So the useful question is not "eat less?" but "eat what?". Two studies from the same research group show the mechanism nicely. In the first, young women received 350-calorie breakfasts containing either 13 or 35 grams of protein. Only the high-protein breakfast lowered the hunger hormone ghrelin across the day, raised the satiety hormone PYY and cut evening snacking substantially - from around 656 down to 486 calories, mostly on high-fat snacks [8]. In the follow-up over twelve weeks, the high-protein breakfast group also reported less daily hunger and gained no body fat [9].

Behind this sits a regularity physiologists call protein leverage: your body defends its protein intake. In the well-known comparison diets, the absolute grams of protein stay remarkably stable - what fluctuates in calories is mostly carbohydrate and fat. In the large clinical trial on ultra-processed food, fat and carbohydrate intake rose clearly while protein stayed nearly constant [10]. Translated: a meal without enough protein is not food to your body. It is an invitation to come back and order again later.

In practice: bring every main meal to roughly 25 to 35 grams of protein. Breakfast is the lever with the longest reach, because it pre-structures your afternoon. If you already eat breakfast, top it up with protein - eggs, quark, skyr or a shake. If you deliberately skip breakfast, that is not a disaster either. Just move the protein to lunch, so the 15:30 gap stays small.

That is why a ProShape All-in-1 Wild Berries lives in my desk drawer these days: on the day when breakfast falls through, the afternoon otherwise ends exactly where the vending machine is waiting.

What makes you overeat: pace, ultra-processing, thirst

That ultra-processed food is hard to stop eating is no longer an opinion, it is measurable. In a clinical trial at the NIH, twenty adults ate ad libitum from ultra-processed meals for two weeks, then ad libitum from unprocessed meals for two weeks - calories, macronutrients, fibre and sugar content of the food on offer were matched to the gram. Even so, participants ate around 508 more calories a day on the ultra-processed diet and gained about 0.9 kg over those two weeks; on the unprocessed diet they lost the same amount [10]. A large umbrella review in the BMJ adds the long view: higher ultra-processed food intake goes together with higher risks of overweight and obesity [11]. Ultra-processed food is not "poison" - but it is calorie-invisible: barely any chewing resistance, and it disappears faster than your satiety system can respond.

Your pace matters too. A review of 22 studies compared fast and slow eating under controlled conditions: slower eaters took in fewer calories - with no difference in how hungry they felt afterwards [12]. That is not an invitation to meditative chewing. It is enough to eat with cutlery, sit at a table and set the fork down between bites. Food shovelled down over the sink counts twice against you: no satiety, full calories.

And the simplest lever stands in front of the meal. In a randomised trial, adults aged 55 to 75 drank 500 millilitres of water before each main meal - and ate less at that meal, ending up with roughly 2 kg of extra weight loss over twelve weeks [13]. To be fair: the effect did not show up in trials with younger adults. Water does not turn cravings into a miracle cure. But a glass of water before the meal costs nothing, does no harm, and buys you a pause before you eat. Fibre does the same job inside the food itself - we cover that in its own article.

The 10-minute plan for the next wave

The next craving is coming - that is part of normal life. What matters more than preventing it is what you do in its first minutes:

1. A glass of water, 500 millilitres. Actually drink it, not a polite sip. Afterwards you have a pause without having to do anything "against" anything.
2. Stay put for ten minutes without opening the cupboard. A craving behaves like a wave: it builds, crests and flattens out - as long as you do not keep feeding it. Mindfulness research calls this vantage point "decentering": watch the impulse as a passing mental event instead of chasing it. Studies show this perspective lowers cravings and goes with less spontaneous snacking [14].
3. Then decide deliberately. If the wave is gone, it was thirst or boredom. If real hunger remains: eat, and eat properly. A snack of protein plus fruit or nuts keeps you full longer than a bag of crisps.
4. Eat slowly, sitting down. The pace trick from the previous section applies here too - snacks included.
5. Keep no banned-foods list. Nothing is forbidden; some things are just portioned more generously. The rules change in frequency, not in permission.

When cravings show up almost daily

One wave a week is weather. One wave a day is climate - and that is when two underlying levers are worth checking.

First: sleep. Do the arithmetic from the meta-analysis: 385 calories a day adds up to nearly 2,700 calories over a week - roughly a full day's energy budget you slept straight through. If your cravings track your bad nights, that connection is your most direct lever, and it is cheaper than any willpower debate.

Second: meal patterns. Cravings bloom in the gaps between meals. Coffee for breakfast and a protein-free salad at lunch builds the 15:30 situation yourself. Three meals with protein spread across the day produce fewer waves than one continuous willpower fight.

And if you notice the wave follows stressful situations almost every time, take that seriously - not as a source of shame, but as information. The route is then less "more discipline" than "fewer triggers": actually take your breaks, defuse the stress sources, and if that alone does not move it, it is worth a conversation with your GP. Eating as a pressure valve is a big, honest topic - and it deserves better tools than shame.

The takeaway

Cravings are not a willpower failure, they are physiology with bad PR. Sleep loss makes you hungrier, stress steers you towards sweets, bans turn every slip into a freefall. Protein in every meal, slower eating, a glass of water beforehand and ten minutes of patience break the wave more reliably than any diet rule.

My default against the afternoon wave stays water, ten patient minutes and protein on the plate. And for the day when a meal drops out entirely, there is ProShape All-in-1 Wild Berries - it fills you up instead of just tasting sweet.

Photo: Trostle, CC0.

Sources: [1] Spiegel et al., Annals of Internal Medicine 2004; [2] Al Khatib et al., European Journal of Clinical Nutrition 2017; [3] Epel et al., Psychoneuroendocrinology 2001; [4] Adam & Epel, Physiology & Behavior 2007; [5] Westenhoefer, Appetite 1994; [6] Paddon-Jones et al., American Journal of Clinical Nutrition 2008; [7] Westerterp-Plantenga, Current Opinion in Clinical Nutrition and Metabolic Care 2003; [8] Leidy et al., American Journal of Clinical Nutrition 2013; [9] Leidy et al., Obesity 2015; [10] Hall et al., Cell Metabolism 2019; [11] Lane et al., BMJ 2024; [12] Robinson et al., American Journal of Clinical Nutrition 2014; [13] Dennis et al., Obesity 2010; [14] Papies et al., Current Opinion in Psychology 2017.