Do night shifts really 'give you diabetes'?

"Shift workers getting too little sleep at the wrong time of day may be increasing their risk of diabetes and obesity," according to the BBC, which reported new research showing that changes to normal sleep can cause the body to struggle with controlling its sugar levels.

The news is based on a lab-based study that examined how three weeks of sleep disruption affected people’s metabolism and blood sugar levels. To do this, researchers recruited 24 healthy adults to stay in a sealed hospital unit for 39 days while lighting levels, temperature and feeding times were manipulated to confuse their body clocks.

At the same time, researchers restricted the number of hours that participants slept each night. They then measured blood sugar levels and metabolism, to determine how the disrupted schedule might affect the body’s ability to process energy.

They found that during the disrupted sleep schedule, the participants’ metabolism slowed and the amount of sugar circulating in their blood after a meal was increased. They concluded that such changes to metabolism could result in an increased risk of obesity and diabetes.

This unusual study provides interesting clues to how disrupted sleep could affect our metabolism. However, the results should be interpreted cautiously given that it was a small, highly controlled study looking at short-term biological changes rather than long-term conditions.

In short, unless you do your job sealed in a small, windowless room for weeks at a time, the study is unlikely to reflect your work environment, and even then it would not necessarily show that your raised blood sugar would lead to the development of obesity or diabetes in the long term.

Where did the story come from?

The study was carried out by researchers from Brigham and Women’s Hospital and Harvard Medical School in the US, and was funded by the US National Institutes of Medicine and National Space Biomedical Research Institute.

The study was published in the peer-reviewed journal Science Translational Medicine.

This research was covered appropriately by the media, with the BBC emphasising that the study’s results should be interpreted with caution, not least as there was a relatively small number of participants involved. Furthermore, the experimental conditions were not equivalent to the conditions that shift workers face in the real world.

What kind of research was this?

This was a before-and-after study in humans that explored whether people’s ability to regulate blood sugar was affected by prolonged sleep restriction and the disruption of their “circadian rhythms”. Circadian rhythms refer to the body’s internal clock, which governs the timing of many factors such as the release of hormones.

The human circadian rhythms work on a 24-hour cycle but can be disrupted by external factors, such as changes in light and temperature. Circadian rhythms can be reset to match these external changes, although some period of adjustment is necessary (this is why jet lag occurs when travelling to a different time zone). Several biological functions exhibit circadian rhythms, including body temperature, our metabolism and the secretion of many hormones. Previous research has suggested that getting too little sleep and disrupting circadian rhythms is associated with an increased risk for chronic conditions such as metabolic syndrome and diabetes.

Human studies in highly controlled laboratory environments have the advantage of ensuring that any effect seen is most likely due to the manipulated variable, in this case sleep duration and circadian rhythm disruption. However, given the artificial setting it can be difficult to tell whether the results of such studies represent what happens in wider populations and reflect people’s real world experiences.

What did the research involve?

The researchers recruited 24 healthy individuals to participate in the study. The participants stayed in individual laboratory suites in a hospital unit for 39 days (approximately 5.5 weeks) while the researchers controlled the unit’s environment. The suites were kept dimly lit with no clocks. The study contained three phases:

  • an initial (or “baseline”) phase lasting six days which consisted of 10 to 16 hours in bed every day, with a consistent bedtime and eating schedule
  • a three-week phase of sleep restriction and circadian disruption, during which the participants spent the equivalent of 5.6 hours per day in bed, while the researchers manipulated the timing of their sleep and eating cycles to mimic an extended 28-hour day
  • a circadian “re-entrainment” (recovery) phase, during which a consistent sleep and eating schedule was reintroduced and the participants spent 10 hours a day in bed

During all three phases, the researchers measured the participants’ weight, resting metabolic rate and post-meal blood sugar levels. They compared these outcomes during the sleep restriction-circadian disruption phase with the initial and recovery phases. They then compared the measures obtained during the three-week restricted sleep-circadian disruption phase with those obtained during the six-day baseline phase, to assess the effect of sleep disruption on these functions.

The data analysis comparing metabolic rate and other biochemical markers before and after sleep disruption can be used to estimate the effect of rhythm disruption on these markers. However, it cannot directly tell us whether they trigger the development of obesity or diabetes over time.

What were the basic results?

In total 24 participants were recruited into the study, although three were not included in the data analysis.

The researchers compared various outcomes after three weeks of restricted sleep and disrupted circadian rhythms with those seen in the six-day baseline phase. They found that, after restricted sleep, the participants exhibited:

  • significantly increased blood sugar levels – an 8% increase in blood glucose when fasting (p=0.0019) and a 14% increase in post-breakfast blood glucose (p=0.0004)
  • significantly lower insulin concentrations – a 12% decrease in fasting blood insulin (p=0.0064) and a 27% decrease in peak insulin concentration after breakfast (p<0.0001)
  • significantly lower resting metabolic rate – an 8% average decrease

Among the 21 participants, three exhibited increased blood sugar levels that would indicate “pre-diabetes” (defined as having relatively high blood sugar levels, often seen before someone develops diabetes) after restricted sleep. No participants had such blood sugar concentrations during the baseline phase (10 to 16 hours of sleep).

The researchers found that blood sugar and insulin concentrations returned to baseline levels by the end of the nine-day recovery phase. The participants’ metabolic rate while resting also increased during the recovery phase, returning towards its baseline level but not rebounding fully.

How did the researchers interpret the results?

The researchers say that their results suggest that efforts to reduce the health impact and risk of diabetes in shift workers should focus on “improving sleep duration” and “strategies to minimise circadian disruption”.


Many people find shift work to be mentally and physically draining, but this small before-and-after study has attempted to find out whether it actually causes negative changes in our metabolism, the system the body uses to produce energy from our blood sugar. While it reveals potential mechanisms by which a disrupted sleep cycle may affect metabolism and blood sugar control, it does not show that shift workers’ sleep patterns cause an increased risk of developing obesity or diabetes. This is for several reasons including the artificial setting and structure of the study, which is unlikely to represent even the most arduous and antisocial shift work performed by most.

When discussing their results the researchers said that they have shown a potential mechanism through which sleep restriction and circadian rhythm disruption may be associated with an increased risk for metabolic syndrome and diabetes. They said that the decrease in insulin production during the disrupted sleep phase led to inadequate blood sugar control, and that this may account for the increased risk of diabetes seen in previous studies. They also concluded that the 8% decrease in resting metabolic rate would translate into a 12.5-pound weight gain over a year (assuming there were no changes to eating or exercise habits) and that this potential weight gain could increase the risk of developing diabetes.

There are several limitations to the study that are important to note when trying to interpret the results:

  • This was a small study that included 24 participants, and analysed data from 21 of the 24 participants originally enrolled. Such a small study size makes it difficult to generalise the results to a wider population confidently.
  • This study occurred in a highly controlled, somewhat isolated environment. While the researchers say that the restricted sleep-circadian disruption patterns may be experienced by shift workers, it is unlikely that the conditions mimic real-world experiences. For instance, in the study the lights were kept constantly dim, something that is unlikely to occur in real life. As light is known to affect our circadian rhythms, it is unclear how varying levels of both natural and artificial light would affect metabolism and insulin and glucose concentrations.
  • The restrictions placed on the participants also appear to have removed opportunities for even basic exercise such as walking, which shift workers would have the opportunity to do every day. It is not clear how much the changes were influenced by a lack of activity, which can affect both metabolism and blood sugar levels.
  • While five weeks seems like a long time to spend in a laboratory, it is not long enough to develop obesity or diabetes. The use of proxy measures, such as resting metabolic rate, to determine likely long-term weight gain and possible subsequent diabetes is not ideal.
  • It should be noted that this study did not aim to determine the impact of disrupted sleep patterns on the development of diabetes, but rather explored possible biological mechanisms that may account for an increased risk seen in previous studies. However, the figures regarding 12.5-pound annual weight gain and increased diabetes risk were reported by the media, so it is important to note that this is an extrapolation and was not an outcome measured in the research.

This study provides evidence that a reduction in the number of hours of sleep each night accompanied by disruption to the body’s internal clock can reduce metabolism and insulin concentrations and increase blood sugar concentrations. However, given the highly controlled nature of this study, we cannot confidently say whether these results would occur in everyday life.

NHS Attribution