The myth of the deadly marathon run

Every couple of months I seem to see another article in a major newspaper about how exercise can be a killer if it is a bit more than a moderate amount. The most recent example: “Over 40? Rethink your next triathlon” (Sydney Morning Herald, 2013-06-24).

About a year ago a study in the journal Mayo Clinic Proceedings made some waves and its lead author, James O’Keefe now keeps getting quoted with warnings to avoid doing more than half an hour of exercise per day. No doubt such warnings will comfort a lot of couch potatoes, who do no exercise at all and who will feel confirmed in their reluctance to go out and move their bodies, lest it might kill them. No doubt his study will be good for O’Keefe’s career, as it makes him the man to go to for a quote for this type of article.

The myth of the deadly Marathon run is as old as the first Marathon run itself. The Marathon distance of 42 km / 26 miles commemorates a messenger called Pheidippides who supposedly delivered the news of the Greek victory over the Persians at the Battle of Marathon over that distance to Athens, only to die immediately afterwards. A closer reading of the historical sources however reveals it to be a myth, created from a mashup of several other events, neither of them involving the death of a runner. Herodotos, in his history of the Persian wars, only mentioned Pheidippides running from Athens to Sparta (a distance several times longer than from Marathon to Athens) to request help against the Persians. No mention was made of a run from Marathon to Athens after the battle or a death from running. After the victory at Marathon, the Greek army rapidly marched to Athens to ward off a sea invasion, but this was no run. Only some 500 years after the battle, the modern version of the story was cobbled together:

The first known written account of a run from Marathon to Athens occurs in the works of the Greek writer Plutarch (46โ€“120), in his essay On the Glory of Athens. Plutarch attributes the run to a herald called either Thersippus or Eukles. Lucian, a century later, credits one “Philippides.” It seems likely that in the 500 years between Herodotus’s time and Plutarch’s, the story of Pheidippides had become muddled with that of the Battle of Marathon (particularly the story of the Athenian forces making the march from Marathon to Athens in order to intercept the Persian ships headed there), and some fanciful writer had invented the story of the run from Marathon to Athens.
(Wikipedia article: Pheidippides)

The modern narrative of death by running doesn’t quite match the facts either. While it’s true that the risk of death from a heart attack is higher during a run than while resting, people who regularly exercise have lower mortality than people who don’t.

As a marathon-running friend of mine explains, being a trained runner has the effect of lowering his resting heart rate. Even if his heart beats faster during a marathon, his average year round pulse will be lower, so his heart will wear out less quickly overall.

While the greatest difference in mortality is found between people who don’t exercise at all and people who exercise moderately, even as people exercise more, the negative effects of exercise do not begin to outweigh the positive effects.

There was an interesting reply in Runner’s World (2012-11-28) to O’Keefe’s claims about excessive training, by which he means anything exceeding 30-60 minutes. In the above article he again advises people to run no more than 30 km a week (which would be no more than about 30 minutes a day averaged over a week). One of the major studies quoted by O’Keefe’s study actually found no levelling off of benefits. His advice is based on a misreading of the available data: People who exercised 120 minutes a day actually had even lower mortality than people exercising one hour a day.

Wen and colleagues reply that yes, they do have data — and it doesn’t show what O’Keefe et al. hope:

By 120 min [per day], the hazard ratio for all-cause mortality was around 0ยท55 [which is better than it was for 60 min per day], with even better hazard ratios for cardiovascular diseases… The adverse effects of strenuous exercise for incremental efforts for more than an hour a day did not seem to outweigh the benefits. We were not able to identify an upper limit of physical activity, either moderate or vigorous, above which more harm than good will occur in terms of long-term life expectancy benefits.

This exchange took place before the recent spate of review articles about the dangers of too much exercise was published. And yet the study is still being cited as evidence that doing more than an hour a day of exercise is bad for you. As a subsequent letter to the journal from Michael Bubb of the University of Florida put it, “The interpretation of the data provided in the review by O’Keefe et al is misleading, particularly given the response of the authors of the original data.”

In the end, articles that discourage exercise that is proven to reduce mortality will end up costing lives.

GPS-logging my bike rides

Five weeks ago I started logging my road bike rides, runs and mountain hikes using the GPS in my Google Nexus S Android phone. I use the iMapMyRide app which requires Android 2.1 and later (of course there’s also an iPhone version).

Start the app, a few taps on the screen and it starts recording. You can pause the recording any time, say if you stop for food or rest. When you’re done you can easily upload the complete route with GPS coordinates and timing to the MapMyRIDE.com website. Besides bike rides you can also use the app for hiking, running or walking.

As it records it displays basic map information, so it can be used for simple navigation too, but most of the time I relied on Google Maps for that.

Afterwards you can view the workout on your PC. It will show altitudes along the route, including total gain. It shows average speeds for each km of progress. It calculates how many kcals you used based on the route, your weight and your age.

A calendar view shows all days on which you exercised, with distances for each workout, weekly totals and monthly totals. This can be a powerful tool to keep up a certain level of exercise on a regular basis.

Battery usage

As with many mobile applications, battery life is of concern to users. So far my longest recorded hike was 5 1/2 hours and my longest bike ride was 4 1/2 hours. I have not run out of power yet, but I’ve had the battery low warning pop up on occasion.

There are a few things you can do to optimize power usage. I make sure to disable WiFi and Bluetooth to minimize power usage. If I am in areas without mobile data coverage, such as high on a mountain I switch the phone into “airplane mode”, which will still let it receive GPS data but it won’t download map data (which it can’t anyway without a nearby cell phone tower). Disabling these wireless connections prevents the phone from wasting energy on trying to reconnect.

It makes a big difference how much you use the LCD screen. If you often turn it on to consult the map for a new or unknown route that will eat battery life.

In order not to have to worry too much about that and to be able to record longer and further rides and hikes, I got myself a cheap external Li-ion battery on Amazon Japan, into which I can plug the USB cable of my Android phone for extra power. I paid JPY 2,380 (about $30) including shipping. Its capacity is listed as 5000 mAh and it has two USB output ports, plus one mini-USB input port for recharging. It comes with a USB cable for charging, a short spiral USB output cable and 10 adapters to connect it to different phone models (including the iPhone and iPod). Because of the standard USB ports you can use any existing USB cable that works with your phone. It’s like running your smartphone off power from your computer.

The device is about the weight and size of my phone. It came charged to about 60%. It should take a couple of hours to fully recharge it from empty.

If fully charged it should theoretically provide three complete charges for my mobile phone, which has a 1500 mAh battery inside, thereby quadrupling the length of rides I can record. Most likely, I will run out of energy long before my battery does ๐Ÿ™‚

My longest distances with MapMyRIDE so far:

  • Bike ride: 71 km, 560 m elevation gain
  • Mountain hike: 14 km, 1020 m elevation gain
  • Run: 10 km in Tokyo

A few rough edges

While the iMapMyRIDE+ app feels fairly solid, it will need fixes for a few problems.

The major issue for me is that the app and the website don’t see eye to eye on time zones. For example, if I record a ride at 17:00 (5pm) to 18:00 (6pm) on a Sunday, the recorded workout title will include the correct time. However, if I view that workout on the computer’s web browser, it is shown on the calendar as having been recorded on the following day (Monday). If I check the details, the start and end times are listed as 8am (08:00) and 9am (09:00): Wrong day and off by 9 hours.

Probably not by coincidence my time zone (Japan Standard Time) is 9 hours ahead of UTC. It’s like the app sends up the start and end time in UTC but the website thinks the data is local time. Yet for determining the date it seems to add those 9 hours again, which takes it beyond midnight and Sunday gets turned into Monday.

I can manually correct every single workout from the website, which also fixes that date on the calendar, but then the app displays the wrong time, which I am prepared to simply ignore.

When I enter my height on the website and then view my details on the app, I am 2 cm shorter than I entered, perhaps as the result of my height having been converted from metric to imperial and back to metric with numeric truncation.

I wish the site would support 24 hour clocks, not just AM/PM. I also wish the site would let the metric size to be entered as cm, not just m and cm separately (probably a hangover from code written for feet and inches).

Note to the app developers out there: The world is much bigger than the US and most of it is metric.

UPDATE 2011-12-13:

I have used the Li-ion battery on two weekend bike rides now. One was 93 km, the other 101 km in length. In both cases I first used the phone normally until the remaining charge level was heading towards 20% (after maybe 4 hours), then I hooked it up to the 5000 mAh battery and continued the ride. The longer of the two rides was about 8 hours, including lunch and other breaks. At the end of the 101 km ride the phone battery was back up to 75% charged, while the external battery was down to 1 of 5 LEDs, i.e. close to empty.

I wasn’t as careful to conserve power with the external battery hooked up. My phone is configured to not go into sleep mode while hooked up to a USB cable, unless I manually push the power button. That’s because I also use it for Android application development, where it’s controlled from a PC via the cable. I should really turn that developer mode off on rides to have the screen blank after a minute as usual even when getting external power. Total capacity with the external battery probably at least 10 or 11 hours, more if I put the phone into “airplane mode”, which disables map updates and hence navigation.

My headlight currently consists of a twin white LED light using a pair of CR2032 batteries that I need to replace every now and then. It’s not very bright, especially where there are no street lights. Probably next year I’ll upgrade the front wheel using a Shimano DH-3N72 dynamo hub,

which can provide up to 3W of power while adding very little drag. A 6V AC to USB adapter will allow me to power USB devices like my phone and the headlights from this without ever having to buy disposable batteries or connecting anything to a mains charger.

UPDATE 2012-01-02:

I have had the front wheel of my Bike Friday rebuilt with a Shimano DH-3N80 dynamo hub. The old 105 hub is now a spare while the rim with tube and tyre were reused. Here is the bike in our entrance hall:

Closeup view of the hub with AC power contacts:

I purchased a USB power adapter made by Kuhn Elektronik GmbH in Germany. It weighs 40 g and measures 8 cm by 2.5 cm. It provides a standard USB-A socket which fits standard USB cables such as the one that came with my Google Nexus S:

USB power adapter with Google Nexus S:

UPDATE 2012-03-14:

At the end of January I started using Strava for tracking rides, in addition to MapMyRides (MMR). I stopped using the MMR app because there is no way in MMR to export GPX files with time stamps, so you can not track your speed or performance on any sites besides MMR. They lock in your data. Instead I either record with Strava on my Android 4 Nexus S or with Endomondo on my Android 1.6 Google Ion. That way I can generate GPX files that will upload to Strava, Endomondo, MapMyRide or just about any other site. The automatic competition feature of Strava is superb. MMR’s best features are its calendar view with weekly and monthly statistics and its mapping feature for planning rides. If those were merged with what Strava can do, it would be a terrific GPS cycling app and site.