Cutting carbohydrates in the evening

What is behind it

Glucose tolerance follows a circadian rhythm. In an eight-day laboratory protocol that systematically decouples circadian phase and behavioral cycle, postprandial glucose in the biological evening at 8 p.m. was 17 % higher than in the biological morning at 8 a.m. – regardless of how long someone had been awake or when they had last eaten. The mechanism: an early insulin response 27 % lower, that is, a beta cell working less well in the evening.

In addition, circadian misalignment, that is, reversing the behavioral cycle by twelve hours, raised postprandial glucose by a further 6 % – but then via reduced insulin sensitivity with a 14 % higher late insulin response. So carbohydrates in the evening do indeed meet a less favorable metabolic state. That is the physiological core of the tip.

Where the chain breaks

What does not follow from this: that reducing carbohydrates in the evening leads to lower weight or better long-term values over weeks. This chain has not been measured end to end in any study. The only six-month randomized study that deliberately manipulated the distribution went the other way: 78 police officers with a BMI over 30, six months, a calorie-reduced diet with carbohydrates mainly at dinner compared with a conventional calorie-reduced control diet.

The evening group lost more weight, more waist circumference and more body fat, had lower hunger scores and better values for fasting glucose, mean daily insulin, HOMA-IR, total, LDL and HDL cholesterol, CRP, TNF-α and IL-6. The authors discuss a hormonal explanation via altered daily profiles of leptin and adiponectin. The study was not blinded, the population was very homogeneous, and no independent replication can be found.

And sleep

Here, too, the data say something more precise than the rule suggests. Twelve healthy men in a sleep laboratory ate standardized isocaloric rice meals with a low (GI 50) or high (GI 109) glycemic index, each four hours before bedtime. Sleep onset latency was 9.0 ± 6.2 min after the high-GI meal compared with 17.5 ± 6.2 min after the low-GI meal (p = 0.009). The same meal only one hour before bedtime resulted in 14.6 ± 9.9 min (p = 0.01).

The meal had no influence on any of the other sleep parameters – sleep efficiency, deep sleep, REM. So it was about falling asleep faster with rapidly available carbohydrates eaten a few hours beforehand, not about better sleep quality. The meals consisted of 90.4 % carbohydrates, which is not a realistic evening meal.

What the studies show

The meta-analysis on the morning–evening difference

Eight randomized crossover studies with isocaloric meals of standardized composition in healthy non-shift workers. Primary endpoints were the areas under the postprandial glucose and insulin curves. The glucose response was markedly higher in the evening (SMD 1.30), whereas the insulin response did not differ (SMD 0.19; 95 % CI −0.10 to 0.49; six studies, 90 participants). Claims about evening insulin spikes are therefore not supported.

Six months of carbohydrates in the evening

78 police officers with a BMI over 30, randomized to a calorie-reduced diet with carbohydrates mainly at dinner or a conventional control diet. On days 0, 7, 90 and 180, blood samples and hunger scores were collected every four hours from 8 a.m. to 8 p.m. The evening group did better on weight, waist circumference, body fat, hunger and a long list of metabolic and inflammatory markers. Not blinded, a very specific population, not replicated.

What meal timing does to weight

29 randomized studies with 2,485 participants over at least twelve weeks. An earlier distribution of calories over the day resulted in −1.75 kg (95 % CI −2.37 to −1.13). Important for this tip: what was studied was the distribution of calories, not of carbohydrates, and only four of the 29 studies concerned calorie distribution at all. 22 of the 29 had high concerns about risk of bias.

What the measurement does not show

There is no randomized study that has tested evening carbohydrate reduction against evening carbohydrate intake with an identical daily balance over a longer period for weight, HbA1c or sleep quality. The well-supported statement concerns exclusively the acute course after a single test meal under laboratory conditions – not a clinical endpoint. And in the same meta-analysis, the insulin response did not differ between morning and evening.

The sleep data come from a study with twelve healthy young men without sleep disorders and concern only sleep onset latency. The weight data on meal timing concern the distribution of calories, not of carbohydrates, and come from an evidence base with a predominantly high risk of bias. The Sofer study, in turn, is a single, unblinded result in a very specific population – that is enough neither for a recommendation nor for a counter-recommendation.

How to do it

For the acute blood glucose course, everything speaks for placing the larger amount of carbohydrates in the first half of the day; in the evening, protein, fat and vegetables remain, plus smaller amounts of complex carbohydrates. For anything beyond that, the data do not point in any direction. Anyone who sleeps well in the evening and is not hungry in the morning has no evidence-based reason to change anything. One detail should not be overlooked: the common addition “no fast sugars in the evening” is at odds with the only sleep study, in which precisely rapidly available carbohydrates four hours before bedtime shortened sleep onset latency. Anyone who wears a sensor and sees their evening spikes has a plausible tool in redistribution – with the caveat that the health value of a flatter evening curve has not been measured.

Safety

Reducing carbohydrates in the evening is low-risk for healthy people because it is a redistribution, not a restriction. Two points nevertheless deserve attention. Anyone taking insulin or sulfonylureas changes the basis of the evening dose with a markedly lower-carbohydrate evening meal and risks nocturnal hypoglycemia – that belongs in a physician’s hands. And anyone who cuts carbohydrates in the evening without fitting them in during the day simply eats less; with a high training volume, underweight or during recovery after exercise, that is rather undesirable. The evidence does not apply to shift workers: the meta-analysis explicitly excluded them.

Sources

• de Almeida RS et al., Journal of the American Nutrition Association 2023
• Morris CJ et al., PNAS 2015
• Sofer S et al., Obesity 2011
• Afaghi A et al., American Journal of Clinical Nutrition 2007
• Liu HY et al., JAMA Network Open 2024
• Chellappa SL et al., Science Advances 2021