Coffee as a fasting booster

What is behind it

Caffeine blocks adenosine receptors, and two things follow from this that matter for fasting. First, catecholamine release increases: in a clamp study in 12 healthy volunteers, plasma adrenaline rose fivefold after caffeine (p < 0.0005), noradrenaline (p < 0.02) and blood pressure (p < 0.001) also rose, free fatty acids rose, and insulin sensitivity fell by 15 % (p < 0.05). An adenosine reuptake inhibitor, by contrast, did not change insulin sensitivity.

This explains both: the increased energy expenditure and enhanced fat oxidation under caffeine as well as the acute deterioration in glucose utilization. Second, adrenaline suppresses appetite – the plausible explanation for why some studies see a decrease in energy intake and others do not.

The hunger effect, examined more closely

The key review separates two things that are constantly mixed up in the fasting scene: coffee as a drink and caffeine as a substance. Caffeine consumed 0.5 to 4 hours before a meal can suppress acute energy intake; coffee drunk 3 to 4.5 hours before a meal had minimal influence on food and macronutrient intake. On gastric emptying, appetite hormones and subjective appetite, the authors explicitly describe the evidence as equivocal.

The most relevant single study recreates the fasting scenario: twelve participants, double-blind crossover, four trials – placebo, decaffeinated coffee, caffeine, caffeine with decaffeinated coffee –, each time two servings of 225 ml of drink plus a capsule after a standardized breakfast, and 4.5 hours later free access to a meal. Energy intake was 2,118 ± 663 kJ with placebo, 2,128 ± 739 kJ with decaffeinated coffee, 2,287 ± 649 kJ with caffeine and 2,016 ± 750 kJ with coffee – no significant difference, and appetite sensations, plasma glucose and gastric emptying did not differ either.

Why the studies are so inconsistent

One possible explanation lies in the genetics of caffeine metabolism. A randomized crossover intervention with 18 volunteers and an observational cohort with 284 participants were stratified by the CYP1A2 polymorphism rs762551 – the gene responsible for more than 95 % of caffeine metabolism. In fast metabolizers, coffee consumption led to lower energy and fat intake, and only in them was high coffee consumption associated with lower appetite sensations and a lower BMI.

Another finding points in the opposite direction: in a randomized crossover study in women with overweight or obesity, drinking 240 ml of coffee together with a standardized breakfast, compared with 240 ml of water, led to an increased craving for sweet foods, a higher fructose intake over the course of the day and higher triglyceride levels.

What the studies show

Coffee against morning hunger – the cleanest single study

Double-blind, randomized crossover design with 12 participants (9 women; age 26.3 ± 6.3 years; BMI 22.7 ± 2.2 kg/m²) and four trials. Standardized breakfast with a tracer to determine gastric emptying, then two servings of 225 ml each of the study drink plus a capsule, and 4.5 hours later free access to a meal. Energy intake did not differ significantly between the trials, nor did appetite sensations, plasma glucose and gastric emptying. The study is small but methodologically clean, and its result is clearly negative.

Meta-analysis on insulin sensitivity

Seven randomized controlled trials in healthy people without diabetes in which caffeine was the only variable changed. Caffeine intake lowered the insulin sensitivity index compared with placebo, standardized mean difference −2.06 (95 % CI −2.67 to −1.44; I² = 49 %; p for heterogeneity = 0.06). The authors conclude that caffeine shifts glucose homeostasis toward hyperglycemia in the short term.

The dose-response relationship

24 healthy volunteers, 12 men and 12 women, in four trials in a randomized double-blind crossover: 1, 3 or 5 mg of caffeine per kg of body weight or placebo, one hour before a two-hour oral glucose tolerance test. The regression slopes for the glucose, insulin and C-peptide areas differed significantly from zero (p < 0.05), the thresholds did not – so the effect begins without any recognizable lower limit. Even the lowest dose measurably worsened glucose utilization.

The autophagy paper is a mouse study

Mice drank coffee; 1 to 4 hours later, autophagic flux was increased in the liver, muscle and heart, as shown by increased LC3B lipidation and a lower amount of the autophagy substrate p62/SQSTM1. At the same time, mTORC1 was inhibited and cellular proteins were globally deacetylated. Crucial for translation to humans: decaffeinated coffee worked just as well as caffeinated coffee, which is why the authors suspect the polyphenols, not the caffeine, as the active agent. There is no measurement in humans.

Where the data stop

The boost to autophagy from coffee has never been measured in humans, and the only paper on it does not support the claim even in the mouse model, because decaffeinated coffee worked just as well. The appetite effect is not reliable: the review calls the evidence on appetite hormones, gastric emptying and subjective appetite equivocal, the best-controlled single study found no difference, and in one study the effect even went in the opposite direction. Then there is the size of the studies: the core papers have 12 to 24 participants, the genetics study 18 in the intervention part. Whether coffee affects ketosis has not been studied. And the blanket claim that sweetened drinks trigger an insulin response is not supported in the fasted state: for sucralose there are findings in combination with a glucose load – a single sip containing 48 mg of sucralose 15 minutes before a glucose tolerance test raised serum insulin at the 30, 45 and 180 minute time points without changing glucose values – and with repeated intake: 14 days of sucralose consumption lowered insulin sensitivity by a median of 17.7 % versus 2.8 % in the control group (p = 0.04). For aspartame, a systematic review with meta-analyses of more than 100 experiments in acute crossover studies found only few effects on blood glucose and insulin.

How to do it

What the studies actually used: 2 × 3 mg of caffeine per kg of body weight, 90 minutes before and 30 minutes after exercise – at this dose, energy expenditure and fat oxidation rose significantly. Two servings of 225 ml of coffee two hours apart – at this amount, neither energy intake nor appetite nor gastric emptying changed. And 1, 3 or 5 mg of caffeine per kg of body weight one hour before a glucose tolerance test – even the lowest dose measurably worsened glucose utilization. The official upper limit is 200 mg per single dose and 400 mg per day for healthy adults. For practice, this mainly results in an order: anyone measuring blood glucose, fasting insulin or HOMA-IR during a fast drinks the coffee afterwards, not before. On the question of whether it breaks the fast: black coffee and espresso contain practically no calories; Bulletproof Coffee with butter and MCT oil, by contrast, contains a relevant amount of fat and is thus by definition no longer abstinence from food. There is no study on the effect of fat in coffee on ketosis or autophagy during fasting. And with sweeteners, the cautious reading is not about insulin release in the moment, but about possible effects with regular consumption.

Safety

The quantity limit is set by the authorities and applies regardless of fasting: EFSA considers single doses of up to 200 mg of caffeine and daily amounts of up to 400 mg safe for healthy adults, a daily amount of up to 200 mg for pregnant and breastfeeding women, and 3 mg per kg of body weight for children and adolescents. In the fasted state, three points are added. First, measurement: anyone measuring fasting blood glucose or insulin gets worse values after coffee, and this starts at about 1 mg per kg of body weight. Second, diabetes: a systematic review of nine randomized trials with 134 participants found in type 2 diabetes, after about 200 to 500 mg of caffeine before a glucose load, an increase in blood glucose concentration of 16 to 28 % of the area under the curve and in insulin concentration of 19 to 48 %, with a simultaneous decrease in insulin sensitivity of 14 to 37 %; in gestational diabetes, about 200 mg of caffeine lowered insulin sensitivity by 18 % in a single study. Third, circulation: caffeine increases adrenaline, noradrenaline and blood pressure, and multi-day fasting already involves losses of sodium and water with orthostatic symptoms – the combination has not been studied, the individual components have. Coffee during fasting is not suitable with caffeine intolerance, with heart rhythm disorders under caffeine, with untreated high blood pressure and in pregnancy above 200 mg per day. Anyone who otherwise drinks coffee regularly should not drop it during fasting: caffeine withdrawal is one of the most plausible causes of the classic fasting headache.

Sources

• Schubert MM et al., International Journal of Food Sciences and Nutrition 2017
• Schubert MM et al., Appetite 2014
• Schubert MM et al., Journal of Applied Physiology 2014
• Shi X et al., Nutrition Journal 2016
• Beaudoin MS et al., Applied Physiology, Nutrition, and Metabolism 2013
• Keijzers GB et al., Diabetes Care 2002
• Pietrocola F et al., Cell Cycle 2014
• EFSA, Scientific Opinion on the safety of caffeine, EFSA Journal 2015