Now that’s a terrible title, hey?

I’ve already written about how you shouldn’t label people who are experiencing a mental health issue here.

So why did I do it?

Well - to quote Kevin Spacey’s character ‘John Doe’ in the movie SE7EN - “Sometimes you need to punch people in the face to get their attention”.

So this blog will address what has - apparently - become a fact - that you will gain weight if you don’t sleep well …

WTF?

In the early 2000s I began ‘on-the-job-training’ in an Insomnia Clinic. That is, learning CBT for insomnia under the guidance of Professor Leon Lack.

And around this time there were some studies being published suggesting that people who slept less had a higher Body Mass Index (BMI).

Leon and I had seen hundreds of people with insomnia by the time these studies were coming out.

And we came to the same conclusion.

Insomniacs aren’t fat.

If anything, their BMI was probably below the norm - because to our eyes - insomniacs were skinny.

My own PhD research at the time showed people with insomnia had a higher 24/7 core temperature than people who slept well (Gradisar et al., 2006).

The higher temperature is likely due to greater metabolism when you have insomnia (we wrote about this earlier).

So why is there research showing a link between sleep and BMI?

The Missing Link

The word ‘link’ can be used in place of words like ‘association’, ‘relationship’ and ‘correlation’.

These words suggest that the ‘direction of effect’ can go both ways.

For example - losing sleep will gain weight - is one direction linking these concepts. But you can also gain weight and lose sleep.

More often than not, when a study finds a link between sleep and BMI, it has measured this at one point (eg, given participants a survey asking them about their sleep, weight and height).

When these surveys ‘crunch the data’, they find the less people sleep, the fatter they are (sorry, I feel I might be losing you so I thought I’d punch you in the face again..).

And this is exactly what we found when we reviewed studies testing this link in teenagers …

Matthew ‘Guido’ Guidolin

One of my former Masters research students was a good little Italian boy we called ‘Guido’ (actually, he came up with that nickname).

Guido found that there was over a 90% chance of finding a link between sleep and BMI if researchers used a single survey.

But …

If they tracked teens over time, there was a 0% chance of finding a link (Guidolin & Gradisar, 2012).

Thus, one of the best ways of finding out if “less sleep makes you fat” is to track people over time. If they lose sleep as they age - do they gain weight?

Sleep Torture

Another great way to test if less sleep can lead to weight gain is to lure human beings into a cavernous sleep laboratory and deprive them of a basic need - sleep.

Experimental studies have shown when us sleep researchers deprive people of sleep, there is an increase in ghrelin (an appetite-stimulating hormone) and a decrease in leptin (an appetite-inhibiting hormone) - suggesting the stage is set to eat more food in the evening (Spiegel et al., 2004).

Indeed, when us sleep researchers need that extra assistance to lure people into our lair … I mean lab … we use cookies.

And experimental studies have shown that when humans are deprived of sleep, they eat more - and sometimes the food they choose to eat makes them plump (eg, cookies; De Leon & Hanlon, 2020).

Can Therapy Help You Sleep And Feel Better?

Another way to test the sleep-BMI link is to intervene. That is, if you can help people sleep better - do they lose weight?

For our skinny insomniacs, we don’t want them to lose more weight.

Probably our best targets are truck drivers diagnosed with sleep apnea.

The treatment for sleep apnea is usually CPAP (continuous positive airway pressure - or in other words, a vacuum cleaner strapped to a person’s face and set to ‘blow’). It opens up the airways of someone with sleep apnea so that they get enough air and continuous sleep.

CPAP treatment does the opposite of sleep restriction - the hormones regulating appetite shift to normal levels (Harsch et al., 2003). Treatment also improves other things beside sleep and hormones - for instance, depression symptoms (Gupta et al., 2016)

Likewise, when insomnia therapy shifts a person from ‘insomnia’ to ‘good sleep’, they feel better (ie, reduced symptoms of depression, anxiety and stress; Sweetman et al., 2020).

Now that I’ve hooked you in with that line above, I can now dazzle you with some info about statistical circles (see below).

If you only crunch sleep-and-BMI data - without adding in data about depression - you have a greater chance of finding a link (…and a greater chance of getting published…).

As the figure below shows, the size of the link (green part) shrinks when you add depression data - because depression overlaps with both sleep and BMI. Thus, the size of the sleep-BMI link shrinks when you statistically add other important variables.

And it’s not just depression that overlaps with sleep and BMI. Physical activity is another important one!

Conclusion?

Fear can sometimes motivate us to act. And problems need solutions.

People working in sleep benefit from being an ‘alarmist’ and then providing solutions to sleep problems.

I confess that I have been an alarmist too - but we need to raise the alarm when we are convinced to do so.

I’m completely convinced that depression is a major problem linked with sleep issues. This has been looked at so many different ways that it’s hard to argue against.

But I’m not completely sold on the sleep-BMI link.

Case in point - I went down a rabbit hole this week when reading one of my PhD student’s manuscripts (Meg Pillion). I got stuck on one of her sentences “Consequences of sleep deprivation include being at risk of obesity (Miller et al., 2018)”.

So I investigated this study.

Meg did the right thing. She cited a really really really good reference - a meta-analysis of studies using prospective designs (ie, tracking people over time). A powerful analytical method of studies using one of the best research designs to test the sleep-BMI link.

I then investigated the studies focused on adolescents - and there were 3 studies (see below).

Strangely, there were 2 studies from the Guidolin & Gradisar (2012) review that were missing. These were the 1st study to test this in teens (Calamaro et al., 2010) and our own study (then again, despite submitting it more than a dozen times to journals, we couldn’t get our study published - see the example of one of the nicer rejections below). Both of these studies found no sleep-BMI link.

Adding these 2 studies would see the sleep-BMI-link move closer to zero.

And these studies aren’t even in a sample of people experiencing insomnia.

I remain unconvinced. How about you?

• Prof MG