Why is it that some women have no period, where others, in seemingly the same situation with exercise and energy consumption continue to menstruate regularly?
A possible explanation for this difference was recently offered by researchers in Sweden. Instead of looking at energy balance or availability for a whole day, as is the standard in energetic research, they computed energy balance on an hourly basis.
Before we get into the meat of the paper, it’s results, and my analysis, a few terms that might be helpful to understand:
- energy balance: Total energy intake (kcal) minus total energy expenditure, i.e., resting metabolism + daily living + exercise.
- resting metabolism: calories burned to fuel involuntary processes: pumping blood, fueling brain, building muscle and bone
- glycogen: stores energy in the liver for short term use, ~300 kcal worth (Farenholtz et. al, 2017)
- kcal: short for kilocalorie, measure of energy. In the US we usually say “calorie” instead.
- negative energy balance: more calories expended than consumed; during small deficits, liver glycogen can make up the difference. During larger deficits (e.g., < -300 kcal), fuel is obtained from other body stores, for example, fat and muscle (called catabolism).
Alright – getting back to the paper. The findings were remarkable; the average time with a negative energy balance of < -300 kcal for athletes with regular periods was 17.6 hours (Interquartile range (IQR) 3.9 – 20.8 hours), versus 21.8 hours in athletes with no periods (IQR 17.8 – 22.4). Athletes with no periods were in a catabolic state (negative energy balance) for four more hours per day than athletes with regular periods.
What I found really interesting was a diagram the researchers included, illustrating how hourly energy balance was calculated. This example shows a period of significant energy deficit at night, with no positive energy balance until the middle of the day (presumably lunchtime). Thinking of energy balance on an hourly basis like this rather than simply the amount of energy consumed in a day leads me to wonder if this might be part of the reason some women can be “all in” for 6+ months without period restoration, where others resume cycles within just 6-12 weeks.
When you look at this graph, note that the total time with energy balance less than -300 kcal is only about seven hours (compared to the much higher numbers described for the women in the study).
Within-day energy balance
In reading this paper and the results, it led me to explore this a possible explanation for why some women restore periods after only a month or two of work on recovery (as described in our book, No Period. Now What?), where others might take six months or longer (and of course, there is the usual caveat that there is no one-size-fits-all recovery, so I’m sure this is not true for everyone!).
I wanted to test this by examining some hypothetical within-day-energy-balance graphs. I started by trying to recreate the graph from the paper based on numbers provided for the participants. The information I used was:
- predicted resting metabolism (63 kcal/hour, 16 waking hours)
- predicted sleeping metabolism (57 kcal/hour, 8 hours sleep as described in the paper)
- daily non-exercise activity thermogenesis (NEAT – basically daily living activities, 446 kcal/day = 28 kcal/hr during waking hours)
- diet-induced thermogenesis (DIT – the energy it takes to digest consumed food, ~10% of calories consumed, in the hour after consumption)
- exercise as needed to fit the graph (25 kcal / hr between 7 and 9 am, 500 kcal exercise at 5:30 pm)
- exercise post-oxygen consumption (EPOC, 8.6% of cal burned through exercise, 2/3 in the first hour post-exercise, 1/3 in the second hour)
And here’s my result:
This agrees with the graph from the paper to a large degree, with around 7 hours in the day with energy balance (EB) < 300 kcal, around 5 hours with a positive energy balance, and 12 hours in negative energy balance (between 0 and – 300 kcal) that would in theory be compensated for by glycogen stores.
To accomplish this, I included the following activities during this hypothetical day:
- Sleep: 10 pm – 6 am
- Wake at 6 am
- A bit of extra walking beyond what was encompassed in NEAT, 25 kcal / hr, from 7 am – 9 am
- Breakfast between 9-10 am of 850 kcal (note the large increase in EB at that time)
- Snack of 80 kcal at ~10:30
- A long lunch of 850 kcal between 12:30 – 1:30 pm
- Coffee, 200 kcal, just before 3 pm
- Going to the gym, 500 kcal between 5 – 6 pm (sharp decrease in EB)
- Dinner of 500 kcal between 6 – 7 pm
- Snack of 100 kcal just before bed at 10 pm
- The only way I could explain the increase in EB at the end of the paper’s graph was that the subject got up from bed to have a “midnight snack” of 225 kcal.
- Total caloric intake = 2805 kcal, total expenditure including all factors listed above = 2805 kcal, daily EB = 0.
ENERGY balance in athletes with no period
Now we get to the interesting part. I wanted to see what the same graph would look like for the amenorrheic athletes. The paper specified that the average maximum deficit was -3181 kcal: almost 2500 kcal lower than in the example above! And these women were spending on average only one hour a day with a positive energy balance – so their within-day energy balance would look a bit more like this:
I couldn’t quite make the numbers work to have a max deficit of over 3000 kcal, nor an overall daily deficit of -600 kcal as described in the paper – so this example doesn’t quite match up with the profile described for amenorrheic athletes – but if anything, their graph would look worse than what I have here ( in this graph the max deficit is -1246 kcal at 9am). Note only one hour with a positive energy balance, and ~12 hours with an energy deficit < 300 kcal. Again, this is not as severe as described in the paper, with a < -300 kcal energy deficit for 21.8 hours for amenorrheic athletes!
The major difference here versus the graph above was exercise in the morning prior to eating anything (leading to the decrease to -1246 kcal between 8-9 am), only a slightly larger lunch (850 kcal), and a larger dinner (800 kcal). This caloric increase mostly but not quite offset the higher amount of exercise (985 kcal vs. 575).
Looking at this graph, it’s easy to understand why these athletes lost menstrual cycles. Hormones take fuel to create; these bodies are conserving every bit of fuel they can!
Within-day energy balance during period recovery
Looking at energy balance in this way led me to thinking about within-day energy balance as a potential difference in women who are “all-in” (in these examples, eating 2500 kcal / day with planned walking for exercise expending 250 kcal, and a daily positive energy balance of 219 kcal – this would go toward things like needed repair, bone density increase, added fat storage). So here’s another example graph:
In this example, the woman has recovered. She eats (please do not get hung up on these numbers, I’m including them simply so you can understand how the examples are the same and where they differ) a breakfast of 500 kcal within an hour of waking. She has two snacks of 100 cal between 9-10 and 10-11. She eats a 700 kcal lunch at around 12:30, has a 200 kcal coffee just before 3, then goes for a walk at around 4:30 and expends 250 kcal. Not too long after that she eats dinner (700 kcal), has a 200 kcal snack just before bed, and sleeps through the night :). She does a pretty good job of keeping her energy balance in positive territory throughout the day, and never gets into the < -300 kcal territory.
Here’s a very different example – but note the overall energy intake and energy balance are the same as in the above graph:
This shows the hourly energy balance for a woman who is eating and exercising the same amount as in the previous example; 2500 kcal / day of food and drink, and walking for ~45 minutes, expending 250 kcal. The big differences are that she takes her walk in the morning, before eating breakfast, and doesn’t get around to eating her 500 kcal breakfast until she gets to work at around 9:15. She doesn’t eat anything else until a 500 kcal lunch three hours later. Coffee in the early afternoon, 250 kcal, and then she feels peckish so has a 250 kcal snack at 4:30. Dinner is between 6 – 7 pm, and she’s used to eating a larger meal so has 1000 kcal. This last meal bumps her nicely into positive energy balance territory… but notice how, contrary to the previous example, she is spending almost about 5 hours in the day with an energy balance below -300 kcal, and another 9 hours between 0 and -300 kcal. It’s hard to overcome the big deficit from the morning, from both overnight metabolism and exercise without eating beforehand. Even though the total number of calories in and out, and daily positive energy balance is the same as the previous example, you can imagine how this woman’s body is not feeling nearly as “safe.”
Based on all this, I recommend that when you are working on recovery, you should try to adjust caloric intake to lessen time with an energy deficit throughout the day. I would strongly suggest eating a good breakfast within an hour of getting up in the morning, and certainly before any planned exercise. Consume meals and snacks throughout the day rather than having fewer, larger, meals. Perhaps eat just before going to bed to provide energy for the work your body is doing while sleeping.
What do you think of this theory? If you have recovered your cycles, what does your within-day energy balance looks like? If you have not recovered cycles, does this ring true for you? Are there changes you could make to keep your energy balance in positive territory for more of the day?