Welcome to week three of my three-week series tangentially related to addiction! (You can find week one here, but it’s probably best to ignore week 2 altogether). As promised, this week I’ll be talking about my personal motivation for the series (besides obviously educating everyone about delicious heroin, which is its own justification). As I’ve beaten past the point of death (and will continue beating, thank you very much!), I’m a firm believer that, even if I’ll never be the person I aspire to be, I can at least force myself to be a crude approximation thereof through a combination of PEDs, successive gender reassignment surgeries, and sheer willpower — which is the reason I do stupid things like rigorously adhere to New Year’s resolutions and give things up for Lent, despite a no active belief in a God who cares, which teeters dangerously toward an active belief in no God who cares.
So this week we’ll discuss why I gave up drinking for 40(+) days and why I’ll be returning thereto once those days are up.
The Problem
I didn’t drink alcohol until I turned 21, like a good little boy. (I had maybe 2.5 beers over the year I was 20, and I don’t think I finished a single one. They were all Corona Light. That may have had something to do with it.) The night of my 21st birthday (which, thank the ambivalent God, fell on a Friday) was A Story that may need to be told later, but was my first real interaction with the delicious intoxicant; it was enough of an encounter that for years I shied away from beer and generally drank only on weekends and holidays, and at those typically not overly much; for the most part, unless I had a reason to be drinking, I wouldn’t really drink.
Sometime early last year, though, at a time of great tumult in my life, I started occasionally drinking wine with dinner. Then I developed a taste for scotch, and I got really into IPAs. Then I started to leave work earlier with the specific intention of having wine with dinner. (The leaving work earlier thing was fine — I generally stayed too late before that; it’s the specific intention thing here that is concerning.) Then I went the entire month of October and had at least one drink every day; then my roommate gave up drinking for a month so during that month I decided I would “drink for two.” Through January and February of this year, I probably drank on average 2.5 drinks per day; on a median day that was probably a half bottle of wine (roughly 3 of your standard Earth drinks), but it varied a fair bit.
I should point out that 2.5 drinks per day is not the end of the world, and it’s not like I was getting drunk most nights. I would have two or three glasses of wine over the span of two or three hours, or I would have a glass of scotch after dinner. I generally wasn’t waking up hungover, I generally wasn’t drinking during the day, and you’ll notice that I didn’t say “I wasn’t sober the entire month of October,” rather, I just had at least a drink every night, and for the most part it was one or two — this is the reason I wouldn’t refer to my time not drinking as “sober,” because I was sober most of the time when I was drinking. I should also point out that when my roommate gave up drinking, I had already decided to give it up for Lent, so it was sort of a “better do this now while I still can” thing rather than anything else. (Also, the title of this post comes from his experience during that time, not my own.)
Still, it was pretty clear that I was drinking more than was healthy — a glass of wine a day may be good for you, but drinking a bottle on your couch alone at midnight on a Tuesday is liver failure waiting to happen. Beyond that, I had three immediate concerns about my alcohol use: alcohol inhibits restful sleep (and I love restful sleep); three glasses of red wine or two beers comes out to almost 500 unnecessary calories a night (and I’m overweight — not cripplingly, and perhaps un-noticeably while clothed, but overweight I am — by about 15 pounds); and alcohol negatively affects athletic performance (and, presumably, other performance as well). It seemed like the math was out of my favor, so I figured I’d keep track of some of this once I quit drinking and I’d see how it all stacked up.
How It All Stacked Up
So — how did it all stack up? I’ve compared the last 39 days (March 5 – April 12, inclusive) with the 39 days prior to Lent (January 25 – March 4, inclusive), and I’m pretty surprised at the results.
Sleep
Admittedly, there are several factors at play here besides alcohol, and these data don’t measure the restfulness of the sleep at all; furthermore, the sleep is self-reported, it’s not like I’m hooked up to an EEG. Still, if you assume that the human errors in reporting are independent of alcohol consumption (always a great assumption) and that external factors are randomly distributed through the 78 days in question, the data suggest that I did not sleep more during Lent than before it:
In addition to the T-Test values indicating that I haven’t gotten higher average sleep since Lent began (p-value is just above .1 for the opposite, so I won’t claim that’s a win for drinking), the shapes of the distributions are clearly meaningfully different; I was apparently more likely to get eight or nine hours of sleep while I was drinking that I have been since I stopped.
Anecdotally, this is explainable by two factors: the first is drinking related, which is that in the first two or three weeks that I wasn’t drinking, I had a lot of trouble falling asleep, which I think is directly related to having given up drinking. The second factor is that in the last two or three weeks sunrise has just gotten to the point where it’s happening before my alarms goes off, so I’m waking up a bit earlier.
Still, the data suggest that my sleep while I was drinking is at worst no worse than my sleep while not drinking, and even if it’s not statistically significant, this suggests that my sleep might be better when I drink. I can’t award a full win to drinking, but it seems better than a tie. Score: Drinking: .5, Not Drinking: 0.
Weight
I weigh myself at irregular intervals, usually in the morning, but sometimes during the day or at night; again, there’s mitigating factors here, but if we assume these factors are uniformly randomly distributed through the last 78 days, we should be able to draw conclusions from the data. I had expected to lose a lot of weight while not drinking — somewhere on the order of a pound a week (500 unnecessary calories * 7 days / week = 3500 calories / week; 3500 calories = 1 lb of fat). What happened instead was that I apparently gained weight:
I should add that I know a lot about day-to-day fluctuations and the like, having had to lose tens of pounds in days, but the repeated observations (n>=12) plus the stark difference in means allows us to draw meaningful conclusions from the data. If you’re losing weight generally, you may gain weight from one day to the next, but on average over six weeks you’ll weigh less than you did before. The message here is that I’m not losing any weight.
However, weight isn’t the only aspect we care about — it’s possible to lose fat and gain weight, by putting on muscle (which could be related to hypothesis #3). I have a body fat-measuring scale, and I record that information too:
There’s not really any meaningful change here pre – vs. post-Lent, which means the weight I gained, inasmuch as it could have affected my bodyfat percentage, didn’t, so I probably gained lean / fat weight roughly in proportion to my pre-Lent lean / fat weight.
A couple of interesting anecdotal points here: I substituted soda for alcohol in some cases; I don’t drink soda very much, but especially in the first few weeks of Lent I would have a Coke when friends were drinking or a soda with dinner; that may have helped to offset the gains from not drinking. However, I still think I drank fewer calories in soda than I would have alcohol; I think the key takeaway here is that my diet, irrespective of my drinking, needs to change in order for me to actually lose weight.
I don’t count the two pounds I gained as a win for drinking given the variety of other factors affecting the outcome here; instead, I count this as a huge loss for not drinking, since I was expecting it would make a huge difference here and didn’t. Not Drinking, I’m disappointed in you. I award you no points, and may God have mercy on your soul. Score: Drinking: .5, Not Drinking: 0
Performance
I religiously record my workouts; how long they take, how far I go, etc. What I should have done was re-trace my workouts from the six weeks prior to Lent and compare them to those same workouts during Lent. What I did do was nothing. I worked out five times a week, and some of the workouts were identical to workouts I did pre-Lent, but not meaningfully many. Instead, this category is entirely subjective.
What I expected to happen was that I would feel like superman. I would be able to lift twice as much weight and run twice as far, twice as fast. I would be bright-eyed and bushy-tailed in the morning, and I would work into the later hours of the night, producing at the level of twelve men. I would never be tired or achy or sore; I would have a positive outlook on life.
What happened was nothing changed; I had good days and bad days, days where I did the same workout I did last week and it took me longer, and days where I went farther or it seemed easier than it had before. I stayed late a few nights, but that was generally because I had to, not because I wanted to get anything done. I could still pinpoint the exact moment I ceased being productive every day. I was still tired and achy and sore; what I had thought was a series of extremely mild hangovers had in fact turned out to be the hammer blow of Life calling every morning, demanding I get up and get out of bed, rather than self-inflicted wounds from last night’s fun. This made my outlook on life decidedly more negative.
This category was basically a wash. Score: Drinking: .5, Not Drinking: 0.
Conclusion
Drinking wins, but by less than a point. The real conclusion here is that Not Drinking doesn’t win. Of course, that’s only in the context of the hypotheses I had going into Lent; there are a couple additional takeaways that shift the math into favor of giving up this experiment once Lent is done. (Actually, I’ll be giving it up before Lent is done; apparently the “Sundays don’t count” rule is true; the 40 days of Lent do not include Sundays, although I did not drink on Sundays, so technically my 40th day ends tomorrow night at midnight rather than Easter morning at midnight. To that end, I’ll be drinking again happily on Good Friday — more than 40 days, but not quite to Easter).
The first takeaway is that I enjoy drinking; I made steak one night during Lent, and I prefer to eat steak with a glass of wine. I enjoy drinking with my friends (and I found I really don’t enjoy not drinking while my friends are drinking). It also helps me unwind — call it self-medication if you like; that’s basically what it is (and this harkens back to the premise of the psychodynamic model from the first addiction post). I’m fine with that, as long as it doesn’t become a problem. What I was really worried about was that it was a problem — that I wouldn’t be able to do it, that I would have intense cravings, that it was negatively impacting my life. I’ve already shown that it wasn’t really negatively impacting my life (at least not in the ways I thought), and while I frequently saw people drinking and thought “wouldn’t that be nice?” I never had true cravings. (And I know cravings — having counted down every day for four months until wrestling season ended and I could go to the grocery store and buy everything on a 40-item list and eat them all in one day, I know cravings. Right now I have Sweetwater on my grocery list, and that’s just because we ran out before I stopped drinking.)
The second factor is that it’s surprisingly inconvenient to not drink. You are always DD, which is great, because at least someone is doing it, except that now you have to hang out with drunk people because you have to get them home, and you’re not having nearly as much fun as they are, but you can’t leave. It’s also really awkward to have to tell people you’re not drinking, because it’s immediately inferred that you have A Problem, and people get really tip-toey about it, like you’re going to judge them for drinking. I at least had the Lent excuse, which of course implied that I’m some sort of religious zealot … try dropping that one on a first date. Oh, and by the way, “going out for drinks” is now off the table; if you want to go on a date it’s either dinner or coffee (not that there’s anything wrong with coffee, but I can’t do it except on weekends).
Finally, if you accept the premise that drinking can be a responsible avenue for self-medication, man did I pick the wrong week to quit drinking.
Since I’ve given up drinking, I’ve been confronted with what amount to fairly minor inconveniences: a team dinner where the cocktail menu featured something called “We Made a Blue Drink!” (I love blue drinks) and a cocktail featuring — and I am not making this up — “meat ice,” which is exactly what it sounds like. I’ve been confronted with social engagements made awkward by my lack of drinking: the team dinner, three dates, and an out-of-town wedding. Duke has lost three times; twice in soul-crushing fashion, once to Wake Forest of all teams, and once in the first round of the NCAA tournament. Brian McCann is playing for the Yankees, and the Braves lit the American flag on fire. And to top it all off, my company laid off 20% of its work force, and as an elected pseudoliaison between management and the proletariat, I have had a front row ticket to management’s We Don’t Care About You Show, plus the distinct pleasure of relaying that message to my friends and colleagues.
In short — man, I could use a drink.
They’re both 39 points. I think it’s just that the Lent ones are way more concentrated in the 7-7.5 range, whereas the pre-Lent one has a couple of 8s, 8.5s, and even a few 9s to smooth the curve out on the right. Again, I think with a higher sample size it’d smooth out on the Lent side as well. Although last night I did sleep 7.5 hours, so maybe not…
Not that it changes the conclusion at all, but what kind of T-test are you doing, specifically for the hours of sleep part? I don’t do hypothesis testing, basically ever, but isn’t it usually based on the assumption that the underlying distributions are Gaussian, which from your plots… isn’t the case?
I think it’s not that unlikely that the underlying distributions can be approximated by continuous random normal distributions (or close to normal; I guess you can’t ever have negative sleep but you could sleep for like 18 hours, and admittedly it probably tends to skew right, since I’m more likely to sleep 10 hours than 4 hours). The weird graph comes from sample size and discrete sampling (I estimate my sleep to within the half-hour, usually), rather than the underlying distribution. In particular, the shape of the Lent sleep distribution looks that way because of a pair of 9.5 and 10 hour data points; if you combine the two datasets you get a much smoother plot (admittedly, still skewed). But you’re probably right, if I were doing an in-depth analysis comparing the distributions I’d need to take into account the probability of tail events between the two datasets and come up with some way of estimating the degree to which the sample size was affecting the likelihood of those events and how that would affect the distributions as a whole, etc., not to mention autocorrelation and probably other factors as well (# work days / various other constraints on bed time / awakening time).
I figured the the reported numbers were rounded to half hour increments, so why is the pre-lent curve so much smoother? The lent one has peaks on the half hour, but the pre-lent doesn’t really. Are you taking more than just the 39 pre-lent days for that curve?
Why on earth do I care about this so much… I need to go outside and/or take your recommendation and do some heroin real quick.
Oh man, a “scientific” experiment? I love it! Why the hell am I not constantly running experiments on myself? I should have measured so many things in DC as a control. Drat.
Regarding your minor weight gain, I know the staggering amount of water you drink during the course of the day. Since alcohol is a diuretic (or at least an anti-anti-diuretic [http://science.howstuffworks.com/alcohol9.htm]), and your BMI didn’t change, could you have put on two lbs of water weight?
Not that I care enough for you to even bother addressing this question, but are the sleep hours normalized for holidays/weekend plans? Or do you set an alarm every day, regardless if you have anything to wake up for or not? Note that I have nothing to wake up for any day anymore.
It’s entirely plausible it’s water weight, although my completely made-up understanding of the way ADH works is that diuretics really only affect you if you aren’t hydrated, b/c if you’re hydrated you’re not producing ADH anyway? IDK.
I did zero normalization for holidays / weekends — just straight up numbers. I (typically) only set alarms on weekdays, but occasionally set them on weekends, which could account for some of the differences (e.g., last weekend I had an alarm set on Sunday — who knows how often that happened pre-Lent?).