Heavy caffeine drinkers sleep less but have deeper sleep, study finds

A recent study using UK Biobank and HypnoLaus cohort data found that individuals drinking four or more caffeinated beverages per day tend to have shorter total sleep time compared to those drinking three or fewer caffeinated drinks per day. While statistical models estimated reductions in sleep length varying from 11 to 229 minutes, the researchers noted that the highest estimates were likely unrealistic, with more reliable causal matching pointing to a reduction of about 11 to 13 minutes per night. The paper was published in the Journal of Psychopharmacology.

Caffeine has been part of human diets for centuries. It is now considered the most widely consumed psychoactive substance in the world. Caffeine occurs naturally in foods and beverages such as coffee, tea, and chocolate. It is added to many energy drinks. Because of this, the regular consumption of caffeine is extremely common.

When intentionally consuming caffeine-rich beverages, people generally do it to feel more alert and energetic, to improve concentration, or to cope with stress. They also consume caffeine-rich beverages because they enjoy their taste and smell, and because these drinks are parts of social customs that facilitate social interaction (e.g. drinking coffee or tea). Because caffeine use is so widespread, its possible effects on health have attracted considerable scientific interest.

One particularly important topic related to caffeine consumption is sleep. Sleep supports essential processes ranging from memory and emotional regulation to immune and cardiovascular functioning. However, studies have shown that caffeine can make it harder to fall asleep, that it shortens sleep duration, reduces sleep efficiency, and alters the depth and structure of sleep.

Moreover, studies looking for substances that help people avoid sleepiness and stay awake regularly examine the effects of caffeine. However, much of this evidence comes from controlled laboratory experiments in which regular caffeine users temporarily abstain before receiving caffeine or consume caffeine as an experimental treatment, situations that may differ substantially from everyday consumption.

Free daily newsletter

Study author Benjamin Stucky and his colleagues conducted a Mendelian randomization study with the goal of estimating the causal effect of habitual caffeine intake on sleep quality. A Mendelian randomization study uses naturally occurring genetic differences associated with a studied characteristic to test whether that characteristic is likely to have a causal effect on an outcome. In this case, study authors wanted to estimate the effects of different habitual caffeine consumption levels on both objective and subjective sleep characteristics recorded at home.

The study authors used data from the UK Biobank and the HypnoLaus cohort. The UK Biobank is a very large open access prospective study conducted in the UK involving a total of 485,511 participants. The HypnoLaus cohort is a population-based study conducted in Lausanne, Switzerland. Its dataset contains data on caffeine intake, genetic data, and objective sleep characteristics data collected using at-home polysomnography (a comprehensive sleep study that records brain waves, oxygen levels, heart rate, and breathing) for 1,702 participants. The polysomnography data were collected during one night.

The study authors used UK Biobank data to estimate the association between various gene variants and caffeine intake, and the HypnoLaus data to estimate the association between each gene variant and objective and subjective measures of sleep quality. Aside from genetic data, study authors used data on the consumption of caffeinated beverages, expressed as the number of cups per day, and data on sleep characteristics.

The sleep characteristics included total sleep time, sleep latency (the time between lights-off and N2 sleep, the second stage of non-rapid-eye-movement sleep, which is a light-to-moderate stage characterized by slower brain activity), number of awakenings during the night, percentage of REM sleep per total sleep time, and some characteristics of EEG (electroencephalogram) activity during sleep, which measures the brain’s electrical signals. Participants also completed self-report assessments of sleep quality (the Pittsburgh Sleep Quality Index), daytime sleepiness (the Epworth Sleepiness Scale) and morningness-eveningness (the Morningness-Eveningness Questionnaire).

The results showed that participants’ self-rated sleep quality and morningness-eveningness (i.e., whether they are an evening or a morning person) did not depend on the consumption of caffeinated beverages. However, statistical models revealed that four or more caffeinated beverages consumed per day shorten total sleep time compared to three or fewer caffeinated drinks per day.

The estimated reduction in sleep length varied from 11 to 229 minutes, though the researchers caution that the higher end of this range is an overestimation caused by statistical uncertainty in the genetic modeling. Interestingly, the shorter sleep in high habitual caffeine consumers was characterized by increased non-rapid-eye-movement sleep depth (as indicated by a specific pattern of electrical brain activity known as delta power).

“The data show that high habitual caffeine intake alters the characteristics of sleep in the general population, while sparing the major physiological principles of sleep-wake regulation possibly due to adaptation,” the study authors concluded. In other words, while people slept less, their bodies compensated by sleeping more deeply, maintaining the natural homeostatic balance of sleep.

The study contributes to the scientific understanding of the likely effects of caffeine on sleep quality. However, it should be noted that the key piece of data the study was based on – the number of cups of caffeinated beverages consumed per day – was self-reported, leaving room for recall bias and estimation errors to have affected the results. Additionally, the study did not take into account the type of caffeinated beverage consumed, even though different beverages can contain substantially different amounts of caffeine. It also did not consider dietary sources of caffeine that are not beverages.

The paper, “Community-based causal evidence that high habitual caffeine consumption alters distinct polysomnography-derived sleep variables,” was authored by Benjamin Stucky, Leonard Henckel, Marloes H. Maathuis, José Haba-Rubio, Pedro Marques-Vidal, Francesca Siclari, Raphaël Heinzer, and Hans-Peter Landolt.