The water treatment industrial complex

Recently bio-nerd, fast guy, and friend of B&P Ethan Linck published an essay in Slate calling into question the empirical reasons behind blanket water filtration in the backcountry, writing: “The idea that most wilderness water sources are inherently unsafe is baseless dogma, unsupported by any epidemiological evidence.”  Slightly more recently, clickbaiting mugwump hack Wes Siler published a rebuttal to Linck’s article in Outside, arguing that “The outdoor community made filtration a must for a reason.”  As he has been for a number of years, Siler makes a compelling argument for how simple and intentional mis(re)statements can drive traffic while at the same time doing the broader argument a disservice.

Linck is clearly (the above quotation is the last sentence in the first paragraph) referring exclusively to wilderness or backcountry water sources, and a review of his citations (especially this one) shows that his thesis is quite air tight.  Siler’s rebuttal, strongest when he contends that the evidence against water filtration is somewhat thin, relies largely on his, one has to assume intentional, conflation of what Linck has to say about backcountry water with frontcountry camping.  I am happy to admit that one undergrad meta-analysis, which is 18 years old and drew on a small body of research, is not the most compelling source.  But in the end Siler can’t argue against Lincks thesis without ceding his point at the same moment; if there isn’t much knowledge proving the safety of backcountry water, there can’t be much knowledge (beyond the anecdotal) proving its danger either.

The ideal solution would be a doctoral-level project collecting and analyzing a broad range of samples from a broad range of backcountry locations.  Stating what agents were present and in what concentrations would be enough information for informed decisions, though speculation on where they might have come from would be welcome.  Getting enough longitudinal data to make such an investigation statistically significant would make it a mammoth undertaking.

While I think ascribing malfeasance to folks like MSR is ridiculous, there can be little question that the folks who make and sell water treatment have little incentive to promulgate such research.  Water treatment has been on the list of essential backcountry gear for at least 50 years.  I can’t recall it not being mentioned in anything published during my lifetime.  It isn’t a subject much mentioned in books written around the turn of the 19th century, and I haven’t read much outdoor literature which spans these two eras.  I can’t recall for example of Colin Fletcher worried about water treatment much while hiking the Grand Canyon, but it’s been a while since I read A Man Who Walked Through Time.  These would be a good historical masters thesis to accompany the above dissertation.

Anecdotally, I’ve had serious GI distress seemingly from unfiltered water only once, when I lazily and knowingly drank half a bottle of water straight from a cow trough before climbing up North Beaver Mesa during the Kokopelli Trail race 11 years ago, but those problems could have easily been from the massive stress I put on my gut during that 20 hour ride.  Tests in two different clinics were both inconclusive.  The effects were severe and highly unpleasant, but passed quickly enough.  Since moving to Montana I’ve become exceedingly lax about treating water, and probably go 4-6 months without doing so on any of my personal trips.  I’ve been more careful with M along, as she prefers as much, and with LB, as that seems the responsible thing to do.  I’m much more rigorous outside the mountains, in placed like the Colorado Plateau and the Missouri River breaks.  Long stagnant water or a river with lots of people and agriculture upstream seem like the right places to treat water, though I have no direct empirical evidence that this is best practice.  When I encounter water like that pictured at top (the North Fork of the Flathead, which is used at its headwaters to benchmark water clarity) I find the notion of tasting it in anything other than it’s original form insulting.

In the end, Siler’s article moves me to name calling for much the same reason I’ve long wanted to push back against the urgency with which water treatment has always been depicted.  Siler starts his article with a cautionary tale, to breed fear and thus caution, as well as doing what I did in the above paragraph, establish personal authority the cheap way.  I was moved to recall the magazine ad ~20 years ( I think from MSR) which depicted a duck pooping midflight into a lake.  The woods are unclean and disorganized, something which on the face of it is something to be managed with water filters and tents and sleeping pads, things which keep us “safe” by mimicking as close as possible the civilized environment which is our default.  The more actively problematic, even pernicious, aspect of this dynamic is the extent to which we humans allow concern to be a substitute for thought.  How we conduct ourselves in the woods has broader import, along with everything else we do, and the way in which water treatment for the backcountry is usually discussed does not reflect well on us* as a culture.

*And by us I mean richish pale people from the global North, for whom things like diptheria are an abstraction when they need to be considered at all.


15 responses to “The water treatment industrial complex”

  1. I’m not quite as daring as you :) but I do drink a lot of untreated water in the backcountry. I’m somewhat selective, but don’t usually hesitate with seeps/springs coming out of rocks/sides of mountains; small, clear tribs; fresh snowmelt, etc. Larger streams/lakes, even more pristine ones, I do usually treat- maybe I shouldn’t be????

  2. First, I am the books for drinking unfiltered/untreated water. Second, while I would not describe myself as an epidemiologist, I did work in an epidemiology department for ten years. The question whether water is safe to drink does not admit a simple yes/no answer in the backcountry. Stuff happens and water flows, so a watering spot that is safe now might not be so tomorrow, and vice versa. Better questions might be, what are the chances I will regret not filtering this water? How much can I manage things going bad if they do? What is ‘bad’ and what is just discomfort? I offer no answer nor advice, yet our discussion of the safety of drinking water (among other wilderness dangers) is forever, in any place and with anybody, coloured by the human inability to have an intuitive feel for probability. That cannot be cured.

    BTW, folk in the 19th century died like flies, so I’d caution against taking their advice on what to do.

  3. I had two complaints about the article:

    1. It relied on 18 year old data for a “high traffic” area of the sierras. 20 years ago “high traffic” in the Sierras was an order or two of magnitude less than it is today. Things may have changed, and relying on data when dozens or maybe a few hundred people passed through a corridor vs thousands during the hiking season matters. And it feels reckless to make that argument and not account for that. The best arguments I’ve seen in response to this is “we really don’t know enough to make a call”.

    1a. The rates of people who are backpacking have, as I mentioned, gone up by an order of magnitude or two. If the argument that most/all of the GI issues hikers get is caused by improper sanitation, you’d expect to see a direct, proportional rate of GI issues increasing as hiker numbers increase, since proper sanitation is a skill. The percentage rate of people who filter their water is very high I’m sure and probably steady. If the numbers go up slowly compared to hiking population, it’d suggest that filtration plays a part in protecting hikers.

    I do not have this data, but it seems like in the absence of wide scale water testing this kind of data could be used to evaluate the issue and get at least a more data-driven interpretation.

    2. The description of the water filtration industry seems… unfair and biased, overly simplified, and relies on an inaccurate portrayal to buttress it’s point. Describing the filtration products out there as all being 100+ dollars and weighing over a pound feels like I was reading an article from 5 years ago. Every water treatment solution I have considered or owned in the last few years cost less than 50 dollars, all but one less than 20-30 bucks, and weigh a few ounces at most. I honestly know of nobody (both personally and in large online communities) who still carries one of those 100 dollar ceramic filter water pumps any more. Misrepresenting the other side of the discussion fast-tracks me towards skepticism and frankly I expect more integrity from someone working towards their PhD.

    At the end of the day, proper hand sanitizing *is* important. But for the majority of backpackers, it’s probably safe to say that they lack the experience and wisdom to be sure of their water sources and what qualifies as back/front country. The solution/insurance against bad water is cheap and reliable and fixes a problem you can’t solve just by looking at it.

  4. You may find this more interesting than the article you referred to: “Backpackers, Don’t Listen To Slate: Science Does Support Stream Water Treatment” at

    (Note: Because of your restrictive comments policy I have to use my Facebook account, which is fake. I don’t care if the comment is published, but please take a look at the story. It’s worth a read.)

    1. I appreciate the link, and for you dealing with the registration I require for comments. I know it adds a layer of complication for some people, but it has also been very effective at eliminating spam and trolls (and makes it easier for me to track and block the IPs of trolls who do make it through).

      That article is certainly worth reading. I would have preferred the author to have less of a knee jerk reaction; the tone, incorrect use of “straw man” throughout, and the preference for quantity over quality with the studies she cites (maybe 1 in 5 are at all relevant to backcountry sources) all make her points weaker. There are some good studies mentioned, if you read most of them you start to get a rather compelling picture of how the presence of stock, and of storms which presumably wash human fecal material into streams, both spike nasty stuff in waterways. That is the sort of the data which is useful.

      What is largely missing from the useful studies mentioned across sources in this discussion is anything beyond the anecdotal association between the presence of ____ in a given waterway and the risk of infection should a human drink ____ amount of said water. As I mentioned, not a simple thing to establish, but most folks seems rather incuriously willing to skip over considering this step.

  5. Warning, meaningless story ahead.

    When we got our dog we started taking him to a dog park with some gnarly water. We tried to keep him from drinking it — no success there. After ~3 months of on-again, off-again frightening mucus diarrhea poop and inexplicable bouts of vomiting he seems to be pretty normal. I strongly suspect he is giardia positive (almost every owner I’ve met at the dog park who bothered to have their dog tested found that to be the case anyway), but seems no worse off for it.

    I would guess that our human systems can adapt pretty well over time to handle almost all backcountry water without issue. But I’ll be honest, at this point, my sphincter isn’t interested in dealing with whatever his sphincter had to. To be fair, I have the ability to be much more selective in my water sources — e.g. I won’t drink from a muddy puddle simultaneously containing dog poop and rotting fish, but for me at least, the answer is pretty easy: filter or treat. If I were away from population centers and on shorter length trips (and not with my wife) I would consider drinking straight from a clean looking source (or at least whatever my subjective definition of `clean’ is). But given that I live near Boulder/Denver its relatively rare that I am somewhere that I don’t think someone has tainted my water source.

    1. Anyone who has experienced or witnessed bad GI issues is easily forgiven any paranoia. I don’t care to change anyone’s habits, but I do think it is handy to consider why your assumptions are what they are.

  6. Your point about nature being something we feel the need to control is really interesting and very well supported in the academic literature, particularly Europe and North America– I can get you a bunch of papers on the topic if you are interested. However, I do wonder how many people view filtering as a way to civilize their time outside (e.g., protect myself from nature while in nature) rather than as a way to ensure that they will be able reenter normal life at will (e.g., not miss work or stick a partner with all the childcare for days because you are sick).

  7. Brave of you to jump into the fray, Dave. And thanks for the discussion, everyone. A few thoughts, aimed at no one in particular, presented in no particular order.

    1) The purpose of review papers in science and medicine the reason is so that the conclusions of disparate studies can be critically evaluated in the framework of peer review to draw general conclusions. Welch 2000, despite its age, remains the only paper to do this. Both the Outside and Discover Mag rebuttals return to the primary literature to grab statistics in support of their argument, which is fine. But in doing so they iron out any discussion of the nuances of what constitutes good evidence, and perpetuate bad habits in science and health reporting.

    What do I mean by that? Here’s an example from the case at hand. We know from a lot of studies that there are variable bacteria loads in backcountry water sources. We know from fewer studies that sometimes these bacteria (and protozoa, and protist) are pathogenic. And we know from even fewer studies yet that these pathogens can be found in concentrations sufficient to make you sick. Similarly, we know from a number of studies that backpackers et al. have high levels of gastrointestinal illness. We know from fewer studies that this illness comes from waterborne pathogens. And we know from very few studies (maybe none?) that clinically-diagnosed illness from waterborne pathogens originated from a backcountry water source where the same pathogens had been assayed for. But for Wilcox and Christie, the only thing that seems to matter is being to demonstrate either the concern or the outcome in isolation.

    And there’s also a whole other issue of how to integrate the multiple papers for quality and for the reach of their conclusions. My editor and I have been responding to corrections requests all week (to which my article remains robust), and people (Wilcox included) keep coming back to a series of papers by Boulware that are essentially variations. At first glance, this seems like a stinging rebuttal to my entire thesis. Look, hikers who filtered were significantly less likely to get diarrhea! But looking at these papers with a critical eye, the conclusions become far more limited. Besides the obvious point that diarrhea itself is not necessarily indicative of waterborne illness (and is defined within his papers in an extraordinarily loose way), there is no attempt to control for the confounding variables in his own studies. For example, in his 2004 Journal of Travel Medicine article “Influence of Hygiene on Gastrointestinal Illness among Wilderness Backpackers,” Boulware *also* finds significantly elevated rates of diarrhea in younger hikers, hikers who don’t take their multivitamins, and hikers with various bad handwashing techniques. It’s pretty easy to see how these things are interrelated and muddy the conclusions. But empirically, it’s a knockout blow to being able to claim much of anything: because the p-value for significant rates of diarrhea of diarrhea in hikers drinking untreated surface water is so close to the standard for a 95% confidence interval (at p=0.03), even a Bonferroni correction for a single compounding variable would bump it into “insignificant.”

    Less relevantly but obviously of interest, 45% of hikers treating their water still got diarrhea per the paper’s definition, which raises questions about its efficacy as an intervention under real-world conditions.

    2) Writing a provocative article for Slate necessarily requires provocative rhetoric. It is emphatically not the same thing as writing a blog post for a popular science magazine, and they serve different purposes in discourse. I think using the example of expensive outdoor gear as an example of the resources we pour into a habit with little empirical support is a reasonable way to grab readers’ attention and does not mean that I am actually mad at Big Water Filter or think they have anything to answer for. Readers seem to disagree, which is perplexing but appears to be my own failure of communication.

    4) What Dave’s post puts nicely and what drives a lot of my interest in the topic – though it’s definitely more of a two-beers-deep argument than fodder for a Slatepitch – is that we as a society have a weird and alienated relationship to the natural world, and finding reasons to fear things in the woods disproportionately to the available evidence of the risk they present is probably not helpful or healthy in that regard.

    5) I think hot takes are good. You might not. That’s also fair.

    6) “Mugwump.”

    – Ethan Linck

    1. “…both Wilcox and *Siler*”

    2. Let me sidestep the issue of water and ask, what is this question: what the hell do people mean with ‘relationship with the natural world’? our rude ancestors were in many ways passively subject to it, in the most brutal way possible — and when the could dominate it, they did so happily and without questions. Both statements are true, mutatis mutandis, today for us, the most powerful critters on the planet. As much as I can, at any time, without any effort, conjure the feel of the tall grass under my hands in a summer dawn in the landscape I feel I belong in, it is all in my head. The grass and the landscape care nothing for me — it cannot be a relationship, because only one party relates. And that experience is there because I am a lucky bastard and it is, in fact, a small thing, with little at stake aside from my own feelings. Taking it further, not really yet fully formed, is the following: is it because what is at stake is not really so pressing (a ruined holiday, some explosive GI experience) that we can stop and ask ourselves ‘what does it mean in our relationship’? how much is too much for us to have the opportunity to ask these questions?

      1. Do we think of ourselves as part of nature, or fundamentally outside of it? What’s our cutoff? Are some cultures inside of it and others out? Do we think of ourselves as animals, as something better, or as something worse? Are ecological relationships necessarily antagonistic, or not? As Dave put it, are the woods something “unclean and disorganized…something to be managed,” or are they home? These questions have widely different answers depending on the context they’re asked in; what I say as an anglo-American in Seattle will be categorically different than Toppy Sundu in Chumbu Province, Papua New Guinea, and Toppy’s answer categorically different than someone in Beijing or a village 70 miles east of Kinshasha, DRC. Nor would we share the same definitions of “dominate” — maybe driving my bigass truck and contributing to the creeping asphalt perimeter of Pugetopolis is dominating nature, maybe it’s not; maybe hunting is dominating nature, maybe it’s just eating or playing. (I doubt our ancestors would agree on these points either.) So sure, consciousness is (maybe, just maybe, with lots of caveats) a uniquely human burden, and only we can really conceive of what it means to “relate” to the biosphere, though this itself will vary in space and time. But we relate regardless, and our philosophical relationships with the nonhuman world manifest as material ones.

        Obviously I dig the question, though.

      2. “The grass and the landscape care nothing for me — it cannot be a relationship, because only one party relates”

        I’ll just add to what Ethan says that I don’t think above can be said simply, either epistemically or metaphysically.

        1. “I’ll just add to what Ethan says that I don’t think above can be said simply, either epistemically or metaphysically.”

          As an atheist, and a materialist I perfectly happy to say you are wrong ;). Partly because I do not believe in metaphysics in any format, and partly because I do not see any nervous system spending any effort in relating to me — here we take that we are talking about ‘relationship’ and not ‘reactions’. I am well aware that there is aways an interaction, and that interaction can very well be started or controlled by something other than myself, but that is like saying gravity is causing a meaningful relationship between me and the planet. A deer might avoid me, but I put this action firmly in the ‘reaction to a stimulus’ box, and I do not see any meaningful two way relationship between the deer and me. Overcoming the deer’s desire to avoid contact might be meaningful to me, but the meaning is fully self generated from an action that involves external agents and entities.

          My take is that our discussion of ‘is there a bacteria in my water’ is purely contingent to our ability to conceptualise the issue. Our rude ancestors did not have a clue, and had been brutally selected to cope with bacterial/viral/parasitic loads that would be a shock to our systems (I’m not saying we would not get to the point of coping, but it might take some painful time before we do). Yet their concerns about the dangers lurking the natural world were no less irrational than ours. I do not see a major difference across the board in terms of ‘how people and culture see the natural world’. We all come from people who left hunting and gathering behind for agriculture. Different people live in different places and thus experience different dangers — that is the contingency that makes people look different in their outlook, but the desire to stay alive, comfortable and disease free are universal. I would also say most people are fundamentally not interested in the natural world if they can at all avoid it. Obviously these contingent differences have practical outcomes, and the simple fact we live in an interconnected world brings people with different attitudes in contact and conflict, which apparently magnify our differences. An example are the different attitudes towards predator management in urban populations vs rural populations. Economically urban populations subsidise rural populations (much to the chagrin of the latter) and thus are entitled to be part of the decision process, causing a major conflict. While the attitudes are different, I see them as being a coherent approach to the natural world: worry about what can get you, otherwise do not. If you do not need to worry, feel free to appreciate.

    3. “But for Wilcox and Siler, the only thing that seems to matter is being to demonstrate either the concern or the outcome in isolation”

      Much appreciated Ethan. You are better equipped than I am to discuss the methodological and statistical issues here, and I appreciate you filling in the things I had to soft pedal over. I think a not-inconsiderable percentage of readers aren’t aware of (and perhaps don’t care) these particular distinctions. You’d think someone writing for a science mag tho….

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s