sleepandcalories_hero_blog (1)

Sound Sleep May Help Curb Calories

Traditional advice on weight management centers around a simple equation — calories in, calories out — although scientists suspect there’s more to it1,2. Calorie counting means logging food (using UP or a partner app) and burning more calories than you consume. While calorie counting works, it’s not much fun. That’s why Jawbone is working to make your food logging go farther.

Scientific evidence suggests that poor sleep can affect the release of hormones (like ghrelin, leptin, and insulin), which regulate hunger and satiety3. In studies that restrict sleep, people eat more calories, particularly from snacks4. What does this mean for UP users? It implies that better sleep might help people feel less hungry or have more willpower to stick to healthy choices.

So, we turned to our rich data to inquire whether UP users who sleep better were more successful with food logging and weight goals. On average, this sample of UP users went to bed at 11:23pm. Normal-weight UP users who went to bed at about 9:30 pm logged 220 fewer calories the next day compared to those who went to bed at 2:30 am. You might expect UP users who are overweight or obese, based on their body mass index (BMI), to log more calories than normal-weight users. In fact, this was only true when their bedtimes fell between about 10:00pm and 2:00am (i.e., based on statistical mean comparisons).



By studying hundreds of thousands of UP users around the world, we identified aspects of sleep that predicted logging significantly fewer calories the next day. We looked at users’ total hours of sleep, average bedtime, how consistently they stuck to this bedtime, and sleep stage. In these analyses, we also took into account basic demographics (age, gender, BMI) and the number of food entries users logged. In general, people tend to underestimate their calories when logging. Nonetheless, the American Diabetes Association’s calorie guidelines recommend that a moderately active woman in her 30’s or 40’s would need to consume about 1500 calories a day in order to lose weight5, and that number is somewhat higher for men. Wondering what these data below actually mean for you? Keep reading – we’ll break it down.

* Bedtime Inconsistency: If your bedtimes vary about one hour (plus or minus) on weekdays, than your bedtime inconsistency is probably close to average.


Next, we looked to see whether early birds (7pm – 11pm bedtimes) tended to eat different kinds of foods the next day, compared to night owls (11pm – 3am bedtimes). As you can see below, early birds tended to eat healthier foods on the whole. This builds on our previous findings that people log more meals high in fats and sugars late in the evening.



Imagine Julia, a 30-year-old woman who wants to lose 10 pounds and logs her meals in the UP App. Julia usually goes to bed at 12am, and sleeps less than 6 hours a night. Sometimes she watches TV until 2am (while snacking), and the next day, exhausted, she collapses in bed at 9pm. With UP reminders, Julia shifts her bedtime to 10:30pm, and becomes more consistent (e.g., 10-11pm most days), and sleeps over 7 hours on average. UP sleep tracking helps her increase the deep, or slow wave sleep, by 20 min a day, and decrease REM by 20 min a day6. What would our best guestimate be for her weight loss based on these data (Fig 2), assuming she made no changes in her activity level? Julia would decrease her calories logged by 100 a day. Assuming that Julia has to log 3500 fewer calories to lose 1 pound5, over the course of the year, this would add up to 10 pounds. These predictions are just a rough estimate, based on our model — there are certainly many other factors we have not captured. However, we hope this paints a picture of what a good night’s sleep might do!


Tucking in early might help UP users eat less and choose healthier foods. A note of caution: these data can reveal links but can’t prove that sleep is the cause. Nonetheless, we know that extra pounds can creep up on us. To reverse the creep, we can make small but consistent lifestyle adjustments — like an earlier bedtime.

What does the future hold for those of us watching our weight? It’s probably not just about calories-in calories-out. At Jawbone, we are actively translating data into personalized insights to help us hack our health.

Technical Notes

To determine which sleep factors were associated with next day calories, we sampled data from hundreds of thousands users and averaged over multiple nights of sleep and next-day calories logged within a user. We constructed a regression model entering all the sleep factors, average number of meal entries logged, and demographics. In this model, total sleep time (duration) was modeled with both the linear and quadratic terms (because scientific research often finds quadratic relationships).

To determine which foods were more frequently eaten by users with early (7pm-11pm) versus late bedtimes (11pm-3am), we sampled millions of meal entries from early birds versus night owls (categorized based on previous night’s bedtime), and limited the data to frequently logged foods. We calculated the ratio of the frequencies with which the food was logged by early birds versus night owls. We computed Chi-square tests to compare normalized frequencies between early birds and night owls, and extracted food categories exhibiting significant differences.

Oat Spectrum_UP2 by Jawbone


1. Hensrud D. Is a calorie always a calorie? (2012). The Mayo Clinic Diet Blog. (link)

2. Aschbacher K, Kornfeld S, Picard M, Puterman E, Havel P, Stanhope K, Lustig RH, Epel E. Chronic stress increases vulnerability to diet-related abdominal fat, oxidative stress, and metabolic risk (2014). Psychoneuroendocrinology, 8(46): 14-22. (link)

3. Morselli L, Leproult R, Balbo M, Spiegel K. Role of sleep duration in the regulation of glucose metabolism and appetite (2010). Best Pract Res Clin Endocrinol Metab, 24(5): 687-702. (link)

4. Nedeltcheva AV, Kilkus JM, Imperial J, Kasza K, Schoeller DA, Penev PD. Sleep curtailment is accompanied by increased intake of calories from snacks (2009). Am J Clin Nutr, 89(1): 126-133. (link)

5. The American Diabetes Association (2015). How many calories do I need? (link)

6. Rutters F, Gonnissen HK, Hursel R, Lemmens SG, Martens EA, Westerterp-Plantenga MS. Distinct associations between energy balance and the sleep characteristics slow wave sleep and rapid eye movement sleep (2012). Int J Obes, 36: 1346-1352. (link)

About The Author

Kirstin Aschbacher

Kirstin is a mashup psychologist-turned-data scientist and mother of a three-year-old who loves to be in movement and learning, preferably at the same time. You can find her peer-reviewed publications on Google Scholar.