Don't know much about biology
August 19, 2010 8:31 AM   Subscribe

Trying to make sense of the readings on my weight scale ...

I'm successfully losing modest amounts of weight, and have observed a loose pattern as I've weighed myself each day. It might be charted roughly like this:

Day 2: down 1.5 pounds
Day 3: no change
Day 4: up .5 pounds
Day 5: no change
Day 6: down 1 pound
Day 7: up .5 pounds
Day 8: up .5 pounds
Day 9: down 1.5 pounds
Day 10: no change
Day 11: up 1 pound
Day 12: down .5 pounds
Day 13: no change
Day 14: down 1 pound

... ad nauseum. Put another way, lose an encouraging amount one day, then a few days of stasis or even gaining back part of the loss, and then another dip. I would love to know whether there is a general biological explanation for the pattern, independent of any factors specific to me, as I could list some of them (changes reflect solely dietary factors, with no exercise; calorie intake varies only slightly each day, with no day where I eat more than the 2700 calories that calculators say I would need to consume to maintain my weight; fluid intake varies only slightly each day, seemingly ruling out water retention as a significant factor; electronic scale) but not all of them. Thanks in advance if anyone can explain a principle that applies.
posted by troywestfield to Health & Fitness (19 answers total) 1 user marked this as a favorite
The only sense I can make of this is that people to do not generally lose or gain weight noticeably from day to day. Your weight on any given day depends more on how hydrated you are, what you ate, and when you last pooped more than it depends on how well you're dieting.
posted by i_am_a_fiesta at 8:34 AM on August 19, 2010 [3 favorites]

fluid intake varies only slightly each day, seemingly ruling out water retention as a significant factor

What about fluctuations in your sodium intake? Sodium (and a few other things related to your diet) could be changing how much water your body retains on a day-to-day basis...
posted by JumpW at 8:36 AM on August 19, 2010 [2 favorites]

Umm... What's the average weight of a bowel movement? 1 pound? Do you always weigh yourself at the same time of day, after you... emptied yourself?
posted by cronholio at 8:37 AM on August 19, 2010 [1 favorite]

I can only guess at the answer to your question (like "is it poop?") but I can give you a solution: calculate a running average. On each day, find the average weight over the previous five days (add the five weights and divide by 5.) If you weigh yourself at the same time each day (ideally right after you get up, after you use the toilet, before you eat or drink anything) and plot this running average I think you will see something much more like your real weight trend.

Some suggest only weighing yourself once a week instead of every day to "solve" this problem but I think that's bad advice.
posted by fritley at 8:38 AM on August 19, 2010 [5 favorites]

It's not so much the fluid intake, but the fluid your body decides to discard and you really cannot measure that (perspiration, urination, etc). The amount of fluid your body discards depends on your physical activity, ambient temperature, what you ate (high in salt? caffeine? alcohol?)...

I once went out on really a hot day and came back a kilo lighter.
posted by xdvesper at 8:39 AM on August 19, 2010 [1 favorite]

Are you weighing yourself at the same time, under the same conditions each day?

I think it's important to realize that it's generally not possible to be as precise about these things as some of us would like. There is no way to calculate exactly how many calories you require, expend, or consume with a very high degree of precision, and certainly not if you're using a simple online calculator. I'm not saying these aren't good tools for establishing a baseline, but the point is that there's variance and getting in the right neighborhood is often the best you can expect. It's not reasonable to think that e.g. it requires exactly 2700 calories to maintain your bodyweight and therefore you can consume exactly 2700 calories and maintain exactly the same bodyweight. There's always going to be some amount of error in your calculation of your caloric requirements or your caloric intake.

That said, the fluctuation in your data looks like random noise to me. I wouldn't worry about it -- as long as the trend goes in the right direction, you're fine -- with the added caveat that bodyweight is a very limited metric to focus on, and body composition, measurements, and athletic performance are often much more useful.
posted by JohnMarston at 8:40 AM on August 19, 2010 [2 favorites]

i_am_a_fiesta has it about right. I've lost 115 lbs since December 2008 and I foolishly did NOT chart my daily weight, instead only keeping track of when I lost a half pound or more. Starting in June this year I have been tracking my weight daily (via True Weight on my iPhone) and the fluctuations are fascinating. I gain at least three pounds during my period. But then I lose that weight plus another pound or two more in the course of the next month.

You cannot get an accurate picture of your weight loss by what the scale says daily, because of hydration, water retention, salt intake, constipation, etc, etc. You need to average it out. I think that's one of the key bits in the Hacker's Diet.

Look into weight averaging charts and keep track that way. As long as there's a downward trend, don't worry about it. I've seen everything from 158 to 151 on the scale this month, but I'm not stressing over it. It all averages out in the end.
posted by elsietheeel at 8:41 AM on August 19, 2010

Yes, to do away with these artifacts of daily life, you should keep a floating average of your weight. If you do that in this case, using a four-day average (because there aren't that many data points), you see that your weight consistently goes down. If you start with the hypothetical weight of 200#, starting from day four on which your weight would be 199, your average weight dips to 197.9 on day fourteen. It plateaus for as many as three days, but it only goes up (very slightly) once (Day 12: 197.9, Day 13: 198, Day 14: 197.9).
posted by OmieWise at 8:46 AM on August 19, 2010 [1 favorite]

fluid intake varies only slightly each day, seemingly ruling out water retention as a significant factor; electronic scale

Are you including the water content of your food as well as beverages? It is not insignificant. (I mention this in addition to the many other potential causes people have already listed, not instead of.)

As for how to interpret the data, The Hacker's Diet recommends calculating an exponentially weighted moving average, a slightly more advanced variation on the 5-day moving average fritley suggests (but if you don't want to get into the heavier math, really, the 5-day average is probably good enough).
posted by DevilsAdvocate at 8:48 AM on August 19, 2010 [1 favorite]

Best answer: I would love to know whether there is a general biological explanation for the pattern, independent of any factors specific to me,

There is. The body's metabolism can be described as a set of non-linear, dynamic processes. Simply put: you are not a balloon and exercise is not like taking air out and eating is not like putting food back end.

Over the long-term such variations will even themselves out, but short term weight gain and losses should be regarded as randomness. This is devilish in the mid-term, say over a week or two time frame when you might see drops of 5 lbs or whatever and think you're losing weight but is really the emergent behavior of several stochastic processes. Sorry if that is buzzwordy.

I did read an empirical study a while ago that simply monitoring your weight 2, 3 times a day (especially before meals) does wonders for weight loss. As you are conscious about your weight, your body responds by suppressing appetite. I did this with no exercise and went from 170 to 155 (probably too low) over the course of 6 months.
posted by geoff. at 8:53 AM on August 19, 2010 [3 favorites]

you should keep a floating average of your weight.

A floating average assumes that underlying model is normally distributed. This is a faulty assumption and if you do this you'll see aforementioned weight gains/losses at the week time scale that will be more prominent in moving the average than they should be. If weight loss/gain were normally distributed you'd see greater variation day-to-day. You don't, the data is much more "quiet" than it should be.
posted by geoff. at 9:00 AM on August 19, 2010 [1 favorite]

I weigh myself almost daily, and this is 100% normal. I've seen my weight fluctuate as much as 3 lbs in 24 hours depending on lots of factors - how much I ate, how much I drank, how much I sweated, and how much I, um, expelled. Even clothes matter - heavy cargo shorts can weigh 1/2 lbs more than gym shorts (I checked :). What really matters is the trend week by week.

To combat all of this, I weigh myself at more or less the same time in the morning and wearing similar clothes. But even then you'll see fluctuations - better to keep on your diet/exercise regimen and watch the trend over time.
posted by Tehhund at 9:02 AM on August 19, 2010

Err, A floating average assumes that underlying model is normally distributed.

Should read "a floating average works best when the underlying model is normally distributed."
posted by geoff. at 9:02 AM on August 19, 2010

I'm not sure I understand your objection to a floating average. While it may not be perfect, it does seem to smooth the abrupt day to day fluctuations and present a more valid picture. I would suggest something larger than a four day average, but to illustrate in this case with limited data it seemed appropriate.
posted by OmieWise at 9:16 AM on August 19, 2010

I personally don't see the point in weighing yourself daily - if you're aiming for the recommended 1-2lbs per week then on a daily basis, even if you could keep everything 100% consistent you're still going to need incredibly accurate scales to notice the 2-4 ounces change. My weight can vary as much as 7lbs depending on hydration, stomach contents and even time of day (seriously, I've seen my weight go up on the scales even if I haven't eaten or drunk anything since in the hours since I weighed myself)
posted by missmagenta at 9:20 AM on August 19, 2010

I agree that weight fluctuates all the time and have found tracking a weekly gain/loss average rather than fussing over daily changes is best to really know what is gooing on (though, like you, I do weigh myself each day, more out of habit than for any good reason).

Also, I wouldn't put too much stock in what the calculators say when it comes to how many calories you can take in to stay at the same weight. Everyone's metabolism is slightly different, as are our activity levels, thyroid functioning, etc. After a while you can determine if that estimate is high/low for you personally by tracking your caloric intake. For instance, I "should be" eating 1800 just to maintain my weight, but really have to stay at 1500 to keep stable with my hypothyroidism, even factoring in exercising.
posted by misha at 9:23 AM on August 19, 2010

It is natural to fluctuate in weight daily, for all kinds of reasons. You just need to track it over time to see which way you are progressing. Graph away, my friend!
posted by modernnomad at 9:42 AM on August 19, 2010

You might also consider that your standard bathroom scale might not be as accurate as you want it to be. Ours supposedly measures to the tenth of a pound. We tried to chart our baby's weight between doctor visits by holding him while standing on the bathroom scale, and then, obviously, weighing ourselves without him. The scale would often measure differently by as much as 1.5 lbs depending on who was holding him, my husband or me.
posted by sillymama at 8:34 PM on August 19, 2010

Response by poster: Thanks very much to everyone for responding. I hadn't posted a question in a while, and forgot how knowledgable everyone here is, which necessitates as many specifics as possible, which is counterinstinctual to me generally, as I tend not to be that open. But in response to some of the questions or points that followed my question:

-- I am weighing myself at the same time each day.
-- I am following an ... um, excretion pattern pretty religiously, particularly as regards weighing in.
-- Level of exercise is also really the same daily.
-- I hadn't thought about sodium intake or, I'm embarrassed to say, the water content of food.

For those recommending a floating average, I should clarify that since observing the pattern, I am not worried, just interested. Knowing the pattern exists makes it easier to accept that I'm going to give back a bit of the progress; I was just curious what reasons might exist for that pattern.

Thanks again.
posted by troywestfield at 8:59 AM on August 27, 2010

« Older Gifts for my 20-something world-traveling...   |   Best writers about web design? Newer »
This thread is closed to new comments.