#1 can't be it, since most obese people probably don't jog. I would imagine the second is very powerful--and also somewhat related to quitting smoking, since the average weight gain after quitting is six to eight pounds, and one in ten smokers appears to gain up to thirty pounds. Smoking is much, much worse for your health than being fat. I imagine #3 and #4 contribute as well.#1: More people are exercising than they used to. How many joggers and walkers did you see on the streets 20, 30 years ago?
#2: Fewer people are smoking. Forget lung cancer (if you can). The big risk for smokers is premature vascular disease. Normally we all have carbon monoxide in our blood (it comes from the breakdown of hemoglobin). [ Brit. Med. J. vol. 296 pp. 78 - 79 '88 ] Natural carbon monoxide production would lead to a carboxyhemoglobin level of .4 - .7%, but normal levels in nonsmokers in urban areas are 1 - 2%. Cigarette smoke contains 4% carbon monoxide, so smokers have levels of 5 - 6%. This can't be good for their blood vessels.
#3: Doctors know more than they did. My brother is a very competent internist. He took over the practice of a similarly competent internist after his very untimely many death years ago. Naturally he got all the medical records on the patients. He found letters (now over 25 years old) from the late MD to his patients informing them of their lab results, and assuring them that their cholesterol was just fine at 250 mg%.
#4: The drugs are better. In addition they may be working in ways that we have yet to fathom. Consider the statins -- their effect on vascular disease is far greater than their effect on blood lipids (cholesterol, triglyerides) -- particularly when compared to other agents that lower blood lipids to the same extent.
(H/t Derek Lowe)
I see chubby joggers all the time. Not morbidly obese joggers - but people carrying an extra 20-30 pounds, yeah. I see people at the gym, who work out regularly, who are really working out, not listlessly pedaling a bike while they read a magazine, who are heavy. Their regular workouts don't seem to be doing much.
The more I learn about how little scientific evidence there actually is that exercise adds anything to weight loss, the more anecdotal evidence I see all around me that it's true.
Yeah, but people who do work out and have what people call "an extra 20-30 pounds" now don't have any higher mortality. They also generally aren't called "obese" at that point, only overweight, and Megan's talking about the obesity numbers.
The thing is that not just "life expectancy is rising," but that the big gain in life expectancy is from a decrease in heart disease, generally thought to be the thing that obesity would contribute most to.
The difference between the top end of "normal" BMI and the low end of "obese" BMI is actually about 30 lbs. One of the interesting things about BMI is that normal and overweight categories span about 30 lbs each, while the obese category spans about 50 lbs. The physique most people associate with obesity is actually that of morbid obesity, while many people would visually classify the obese as merely overweight.
So yes, many joggers with "an extra 30 lbs" could easily be obese. I see plenty of people exercising in my area who most likely qualify.
Their regular workouts don't seem to be doing much.
Eh, it's entirely possible that losing weight is only one of their goals. I worked out pretty consistently (2-3 times/week) pretty consistently during the years when I was 20 lbs overweight. I kept going despite not losing any weight because of a noticeable improvement in fitness. It lessened the blow of being the fattie in the room if I was one of the few people not huffing and puffing when my team at work had to walk up three flights of stairs to the alternate conference room.
I actually lost the weight when I moved my workouts to 4-5 times/week.
I would definitely agree with #2 through #4. However, I think they can all be summed up as the following: "we know more about what kills us than we used to, and are developing new ways to treat/prevent the things that kill us every day."
I know we like to point to the high punitive taxes on tobacco as a primary reason for the decreasing prevalance of smokers, but I think younger generations are much more aware of the disastrous effects smoking has on your health - which has helped socially stigmatize smoking in a way previous generations were unable to stigmatize it.
I don't think obesity can be correlated to life expepctancy without adjusting for so many other health factors, including comorbidities such as vascular disease. I would not be surprised if obese indviduals without hypertension or vascular disease outlive non-obse indviduals with such disease. It's what lies beneath that kills you...
which has helped socially stigmatize smoking in a way previous generations were unable to stigmatize it.
I agree, how many people allow smoking in their house anymore? Nearly all smokers I know go outside to have a smoke, and zero non-smokers allow smoking in their house or cars.
Just as an example of how things change, when my parents stopped allowing smoking in their house in the 1970's family members threatened to not come over and were constantly caught trying to sneak one inside.
No. 1 surely isn't the whole story, but it can be part of it even though obese people don't jog. We're dealing with averages here, so if fit people's life expectancy goes up because of exercise, that helps the averages, if only by offsetting part of the decline caused by the health problems of the sedentary.
I knew a guy once whose theory of effects was "nothing matters," by which he really meant that no one thing ever matters enough to generate big measurements. Not absolutely true, but sometimes pretty close.
I put on 10 lbs in one month after I quit smoking (thank Pfizer for Chantix, btw.)
I weigh 60 lbs more than when I left the Army, but still walk 3 miles/day at a 15 minute mile pace. Exercise alone doesn't help with weight loss, you have to eat less too (chocolate chip cookies will be my downfall...) But exercising more+eating less works better than just eating less.
Chantix is a great medicinal chemistry story -- it's really worth trumpeting in terms of basic pharmaceutical innovation. It doesn't get good press because of the rather serious side effects that some people have, though.
No1 seems pretty annecdotal. It's also probably offset by the fact that fewer and fewer people are engaged in jobs that require manual labor.
Yes, the model is wrong. First, the life expectancy is directly based on the death rate. The death rate has decreased a lot, and the paragraph below is from the CDC:
http://www.cdc.gov/nchs/data/hus/hus08.pdf
In 2005, a total of 2.4 million deaths were reported in the United States (Table 30). The overall age-adjusted death rate was 45% lower in 2005 than in 1950. The reduction in overall mortality since 1950 was driven mostly by declines in mortality for such leading causes of death as heart disease, stroke, and unintentional injuries (Figure 16).
In 2005, the age-adjusted death rate for heart disease, the leading cause of death, was 64% lower than the rate in 1950 (Table 35). The age-adjusted death rate for stroke, the third leading cause of death, declined 74% since 1950 (Table 36). Heart disease and stroke mortality are associated with risk factors such as high cholesterol, high blood pressure, smoking, and dietary factors. Other important factors include socioeconomic status, obesity, and physical inactivity. Factors contributing to the decline in heart disease and stroke mortality include better control of risk factors, improved access to early detection, and better treatment and care, including new drugs and expanded uses for existing drugs (1).
So although obesity has gone up, early detection and innovative drugs (yes, due to the private drug companies, no less) have more than offset obesity.
Obesity is a risk to increased mortality rate. Its just that advances and innovations in health care, provided for private drug companies in the pursue of profit, more than offset it.
Therefore, yes, the model is false.
I'm obese (BMI 32), and I jog 3-4 times a week.
Please don't buy into the stereotype that says you can judge people's diets or activity levels based on body size. It's just not true.
I agree with Amy. My BMI is 32 and I run or bike 4-5 times a week. Just got done with my annual physical and my health checked out 100%. The Dr. didn't even hint at losing weight because I'm active and have no other risk factors.
My unscientific opinion is that humans in developed nations are getting larger because of the availability of quality food (especially proteins), and improvements in medical science and physical fitness. "Obesity" is rising because the accepted measurements like BMI haven't been adjusted accordingly. In other words, there's a new normal. I have a friend in the fitness apparel industry that agrees. She tells me that today's size "medium" was size "large" 20 years ago. I heard the same thing for women's apparel (that what used to be a 4 is now a 0).
It's not that people are getting fatter, they're getting larger (and healthier).
Actually quality protein is quite expensive. It's carbohydrates that are really cheap and ubiquitous. People are getting obese on Mac N Cheese, not a steak.
Something seems funny about the math in no.2
One thing I never understood, and perhaps someone here can explain, is why cig smoke contains carbon monoxide to begin with. Normally, when an organic fuel is burned, the reaction is CxHx+O2-->CO2+H20. Carbon dioxide is a much more stable molecule, and it, not CO will form unless there are some extra ordinary circumstances. Engines provide one such circumstance, since the oxygen levels are limited by the valves at the time of combustion. However, there is no limitation of oxygen in a cigarette, since its exposed to air, so what gives?
Natural carbon monoxide production would lead to a carboxyhemoglobin level of .4 - .7%, but normal levels in nonsmokers in urban areas are 1 - 2%. Cigarette smoke contains 4% carbon monoxide, so smokers have levels of 5 - 6%. This can't be good for their blood vessels.
This is the part that confuses me. Does "So smokers have" mean a measured result, or an assumed one? If its measured, fine, they know something I don't. But if its assumed, than it sounds like someone expects smokers to breathe nothing but cigarette smoke 24hrs a day.
Ken,
Perhaps if you think of the shredded tobacco wrapped in paper as the combustion chamber, there are areas within the cigarette where combustion is occurring in a low O2 environment?
I can't explain why cigarette smoke has carbon monoxide, probably everything that burns has it to some extent, but we don't breathe it in, like smokers do. Your lungs, if all the little air cells (alveoli) were spread out would be the size of a tennis court. Carbon monoxide and oxygen bind to the same site in hemoglobin and compete for it. Unfortunately the affinity of carbon monoxide for hemoglobin is 200 - 250 times that of oxygen. So once monoxide binds, it stays there. That's why acute poisoning is so bad -- it's like losing that amount of oxygen carrying capacity of your blood at once. This is also why one of the treatments for monoxide poisoning is the hyperbaric chamber -- which pumps in oxygen at high pressure, so it is better able to compete with the carbon monoxide for the binding site. If you're a chemist it's Le Chatelier's principle in action
Not to go all Masonomics, but jogging/public exercising like that mentioned in #1 is almost certainly about signaling and little else.
I'm one of those 30-40 pound overweight people you see out, down from 60-70 earlier this year.
I've been running 1/2 marathongs for a few years (1 or 2 a year) and occasional 5 and 10k's as well. I run/jog outside because of the distance and nothing about signaling. I can't stand doing hour+ runs on a treadmill. I also bike to work ~10 miles each way in a bike jersey and pants/cargo shorts, since I don't like my wedding tackle being on display.
Side note: I stayed at that 60lb range for years with just the running. I had to add a few days of high intensity classes and some speed work to start taking it off.
The main statistical determinants of differences in life expectancy are smoking and how long you stay in education. Obesity is roughly nowhere by comparison.
However people who are mildly "overweight" tend to live longer with less morbidity than those underweight, normal weight or obese. Maybe part of the trend is a rise in the proportion mildly "overweight"?
Are you saying that smokers have lower education than non smokers, and once this difference is adjusted for life expectancy is the same for both groups?
Educated non-smokers live longer than educated smokers. Nowadays smokers tend to be less educated than non-smokers; but edcation and smoking have distinct, and major, effects on life expectancy.
I've seen this link bandied several times now in the comments section, but it's quite misleading. When a study properly controls for smoking and chronic illness, the conclusion is opposite. Being overweight decreases life expectancy:
http://www.futurepundit.com/archives/003680.html
Thanks for the link. It is rare to see the New England Journal of Medicine correcting the Centre for Disease Control but that is what has happened here. Barring that the 'ideal' Body Mass Index has probably been set a bit low; being overweight does tend to shorten your life.
Since we're looking back over several decades, it may be worth tweeking #2 to include the general drop in pollution and chemicals in our environment. Advances in machinery, construction standards, etc. have us living in a much cleaner world than we did in the 50's to 70's.
Not if you live in Hamilton, Ontario. AKA "the armpit of Ontario". The air is saturated with chemicals and the town as the highest rate of cancers in all of Canada.
You really should to talk to someone with an exercise science or a related degree and ask them why people fail at losing fat and keeping it off. There has been a severe lack of expertise here posted here, while giving lots of credence to a lawyer who wrote a book that shows that people who exercise don't lose weight, but doesn't get into why in significant detail.
Though there are the somewhat rare hard luck genetic cases (Pima Indians), generally speaking, people really are doing something wrong.
Amy's right. Obese people exercise all the time, even jog. Check out the Clydesdales and Athenas. These people are serious, consistant, long time exercisers and these people do it because they enjoy it. To look at them people may think they're 'overweight' or 'could lose a few pounds, or even 30' but would not consider them 'obese', and in many instances can't even think of them as obese simply because they are able to run.
People say 'oh the obese don't exercise', but that's only because what you think of as obese may not be accurate. Non-obese (are you sure?) people think all obese people are like the headless photos that mark every article on weight. These people can hardly move. But there are plenty of people on the low end of the obesity range. If you're a woman who's 5'4 and 175 lbs, you're not even wearing 'plus' sizes, you wear a size 14, a large, maybe even a size 12 depending on your shape. Size fourteen is the American average. You've never seen a size 14 woman jogging?
But those people aren't the ones with the health problems. By the time you're heavy enough to really damage your health, you're also heavy enough that it's really, really hard to jog. Or so I understand.
Now you're moving the goalposts. If obese people at the lower end of the BMI scale can (and many do) exercise regularly, that could explain a good chunk of why obesity and life expectancy are both rising. Most obese people fall on the low end of obese on the BMI scale, and are therefore capable of exercise. So more exercise in that group could explain the trend, even if the "superobese" are not exercising.
For the "average" American woman who stands 5'4", the "morbid obesity" label kicks in at 232 lbs. That's obviously fat, to be sure, but are you seriously arguing that a 230-lb woman is all but incapable of going for a fast walk or a bike ride or a swim?
Not really; look up the BMIs for your professional football players sometime. http://jama.ama-assn.org/cgi/content/full/293/9/1061-a
Yes, but it's generally accepted that BMI is an imperfect measurement; this is in part because it was developed in response to insurance data from the 1930's. It generally *does* overstate obesity in the heavily-muscled subset of the population, for whom waist-to-hip ratios are probably a more accurate number if you insist on it being purely morphometric.
For people who are *not* lifting weight or performing heavy manual labor, BMI is pretty accurate.
your stats are too generalized and subject to too many biases. see the post previously.
Gallup just published a new study that tracks the obesity and diabetes across states. the correlation between obesity and incidence of diabetes is 0.6, which is very strong.
This is a direct result of obesity means more disease. it is common sense after all.
life expectancy vs obesity is affected by exogenous factors such as drugs, better detection and diagnosis to show a strong correlation.
http://www.gallup.com/poll/122405/Obesity-Diabetes-Across-States-Clear-Relationship.aspx
I actually see far fewer joggers than I did in the late 70s and 80s - by a large factor. It's especially true for young adults and women. Most of the runners I see tend to be men in their 40s. I cannot tell you when I last saw a woman or 20 something running.
I live is a safe, quiet suburban area and walk my dog a lot so I tend to see the goings on.
Really? That doesn't match my experience at all. Female joggers outnumber males about 3:1 where I run. Though it might be driven by demographics -- I live in a disproportionately young and childless neighborhood in an active city. My guess is that childless 20-something females are much more likely to go running than married-with-kids late 30-somethings. And the opposite for men.
A lot of them don't run on the street but do it inside a gym. Safer + better on the joints using an elliptical machine.
I'm guessing you could be right on the demos. My area is all families. Singles will do more exercise because they have more time and want to remain attractive to the opposite sex.
I also think that, in my area, there is an exaggerated fear among women of something like being raped or otherwise attacked. It's driven by TV news about stuff in New York City, not here on Long Island.
How many joggers and walkers did you see on the streets 20, 30 years ago?
A lot. More than now, probably...the running boom really took off in the 70's.
I don't know about joggers, but road race participation is at an all-time high. There's a lot more races, and a lot more people in every race. Heck, the Boston marathon sold out last year (difficult to do, since you have to qualify, and only about 10% of marathoners do that).
Statins, antioxidants, and diabetic maintenance.
Some argue that the rise and decline of heart disease reminds them of an epidemic of an infectious disease. So they wonder about some unidentified infectious agent. Before you laugh, remember all the propaganda about ulcers being caused by stress, until an Aussie showed that many of them were, indeed, caused by infection.
The book is a little dated but Ewald's Plague Time is an excellent survey of the idea of infections causing more chronic diseases than is commonly considered.
Yay Paul Ewald! :-)
It's because BMI doesn't measure fat content, just weight vs. height. As people exercise, they build muscle and destroy fat. Muscle weighs more than fat so the fitter you are, the heavier you are.
BMI is basically useless as a measure of health. No surprise then that the government uses it.
I agree. Typically the government is 30 years behind the real nutrition/fitness experts. They'll start using body fat % in 2039. Meanwhile those of "in the know" can laugh at the ignoramuses.
That's because body fat% can't be easily administered at home. I know there are scales, but they are very inaccurate. BMI gives you a very rough idea, and likely works well for large scale population studies. For individuals, you'll need to use some common sense. If you lift weights 6 times a week and have a six pack, a high BMI is due to your muscle and not fat so ignore it. This same argument is brought up in every discussion about BMI, usually about how Michael Jordan is obese or something similar.
We understand the caveat that it isn't useful for everyone, and some common sense needs to be applied. Until there is a reliable way to step on a scale and have body fat% spit out, it won't be a tool that can be used at home.
That's true, it's difficult to determine body fat % on your own even with calipers. However it's enough just to look at your fat naked body in the mirror to see what's the problem. Just go on how you look - the mirror doesn't lie!
#3 is more important and complicated than just "doctor's know more". It's not just a question of medical professionals knowing more what a healthy cholesterol level is. It's that medicine has progressed in such a way as to keep unhealthy people alive longer.
For instance, a man I know had a heart attack at 38, due in large part to poor diet, no exercise, and an extremely high stress lifestyle. After a brief effort, he returned to his poor diet, lack of exercise, and stressful job. He had another heart attack at 49. He is now 60. Though the second heart attack did finally force him to make some changes in diet and exercise, it's unlikely he would have survived both heart attacks without advances in cardio surgery, angioplasties, stints, and drugs. Those advances essentially kept him alive for 15 years until he decided to make changes for quality of life reasons. Oh, and by the way, it's very likely that a prescription for anti-depressants helped him choose to make those changes, as well as lowered his general stress levels. 30 years ago, he'd be dead.
A less happy example is a woman I know who has been battling lung cancer for almost 15 years. She's gone into remission multiple times, but each time the cancer has returned more viciously and spread more rapidly. Unlike the man in the example above, her quality of life has deteriorated over time. But she is still alive against the odds, in large part due to advances in chemotherapy. She is in a lot of pain and exhausted most of the time, but her life has been extended.
A longer average life expectancy does not mean that people are necessarily healthier than we once were. In many cases, we've adapted medicine to keep us alive despite major health problems.
As a fun statistical exercise (I did it in Grad School and don't have immediate access to the datasets, but I am sure you could find them if you looked around), you can do a simple correlation of the fraction of population that smokes against the fraction of population that is obese.
The correlation coefficient was like .85, i.e. enormous.
Now this is likely the result of obesity increasing and smoking decreasing in our society through time, but it does make you think.
Life expectancy is going up but I would contend that quality of life is not. We have the technology to keep half dead people alive for years on machines.
Obesity is a direct result of diet and you can't outrun nutrition. Grains, low cost carbohydrates, represent as much as 80% of some diets and grains/sugars will make you fat, regardless of exercise. Cut the grains - eat meat and vegetables, nuts and seeds, some fruit, little starch and no sugar. If you clean up your diet, you can work out minimally and still be lean and muscular, extending both life expectancy and quality of life.
www.crossfit.com - spreading the word
Perhaps the optimum level of radioactivity in the environment was approached more nearly courtesy of the nuclear weapons tests of the 50s and 60s. Since then perhaps the use of radiation in medicine has kept the tally up.
One major source of increased obesity that hasn't been mentioned is the widespread use of air conditioning. Growing up in the Northeast during the 1950s and early 1960s, none of my friends had A/C in their car or house. Today we think of the South as the center of the obesity epidemic, but my recollection was that it was very different back in those days.
Nothing like having temperature in the mid-90s to reduce the temptation of the all you can eat buffet.
Nothing like having temperature in the mid-90s to reduce the temptation to go out and run a few miles, either.
Lifelong southerner here, in an area that was moderately affected by Katrina. My power was only out for a couple days, but many of my friends were without AC for up to two weeks. I can promise you, all any of us wanted to do was lie around in the shade -- nobody had the slightest inclination to go out and voluntarily sweat even more.
#2. I've never looked into this, but my cursosry understanding was that, unless Carbon Monoxide causes asphyxiation it doesn't cause any lasting harm. It may complex with the hemoglobin to form carboxyhemoglobin, but I don't know that it damages vessel walls. I think the damage from smoking comes from elsehwhere.
Re: Cholesterol, it's worth noting that European docs are fine with higher cholesterol levels than Americans are.
Emergency care has improved remarkably; the ambulance is more likely to bring you to the hospital alive (which also accounts for some of the change in homocide rates, I believe.)
How do the rates of heterocide and homocide compare?
I don't believe there is anyting wrong with the model. It could be as simple as fat house cats vs. stray outdoor cats. A cat will live longer in a house, and most likely be fatter, depends on how the owner feed it. The stray cats will been skinny depend on where they get their food from but im sure other stress factors can shorten the live of the cat.
So we are all a little fatter, so maybe that will trim a few years off. But we have got better health care, and more safety measures. If we saving a 5yrs old, than a few 60yrs olds, I think that as a better for the life expectancy.
Re: "I imagine #3 and #4 contribute as well."
Is that what The Atlantic is paying you for? To "imagine" things?
Huh.
Two points:
1. Show me a study that says people who reduce cholesterol from 250 to 190 live longer.
2. The definition of obesity has changed in that time interval, which accounts partially for the larger number of obese people now.
As you requested:
The landmark WOSCOPS trial (n=6595) reduced total cholesterol from an average of 272 to 218 and decreased mortality by 22% (p=0.051). Yes, it barely missed statistical significance, but all of the trends are clear. Cardiovascular mortality reduction was 33% and *was* statistically significant at p=0.033.
All subsequent statin trials (I'm just dealing with primary prevention, not the secondary prevention trials, which were positive as well) showed major clinical benefit in decreasing myocardial infarctions, cardiovascular morbidity, and various combinations thereof. None of them showed a statistically significant decrease in overall mortality, largely because they were only powered for combination endpoints for cost reasons, but all showed trends toward decreased mortality. I suspect that a meta-analysis *would* show a significantly decreased mortality.
Modern statin drugs decrease cardiovascular mortality and morbidity and decrease overall mortality in any population with appreciable cardiovascular risk. That's as well known as just about anything in medicine.
The problem with this sort of thinking, is that you have to accept the bad along with good. Just plot the drop in ambient lead levels over the past 30 - 40 years against time. Now plot the drop in College Board scores the same way (before the Board normed them up so they wouldn't look so bad -- this really happened). The two curves look the same.
Correlation is not causation
Sorry -- the above is really a comment to Tim McGuire's comment, that declines in polution might be responsible for the drop in mortality.
Also to Tony61 -- there are huge numbers of studies showing declines in heart attack and stroke rates when cholesterol levels are lowered. Since both of them can and do kill people, they also show a decline in mortality.
The first was the Scandinavian Simvastatin study [ Lancet vol. 344 pp. 1383 - 1389 '94 ] Simvastatin (Zocor) is a lipid lowering agent thought to work by inhibiting HMG Coenzyme A reductase (a crucial enzymatic step in the biosynthesis of cholesterol). This is the first study showing that giving it to patients with elevated cholesterol (212 to 309 mg%, e.g. over 5.5 milliMoles/liter) and a history either of angina or myocardial infarction actually improves survival.
4444 patients from all over Scandinavia meeting this criterion are reported with 5.4 year median followup. The total cholesterol dropped 25%, the LDL cholesterol dropped 35% and the HDL cholesterol increased by 8%. The mortality in the simvastatin group was 182 patients (8%) vs. 256 in the placebo group (12%). There were 189 coronary deaths in the placebo group and 111 in the simvastatin group.
There have been many more since.
I'm pretty sure those studies were done on those who had already had a heart attack or other heart problems. Not the same as someone who only has what's considered to be "high cholesterol" but is otherwise healthy.
You're right about the trials above, but the primary prevention trials (WOSCOPS, AFCAPT/TexCAPS, ASCOT-LLA, METEOR, and JUPITER) all showed pretty much the same thing. Yes, all of those trials used an additional marker of increased risk (like high CRP, low HDL, etc., etc.), but here is absolutely no question that lowering cholesterol in essentially *any* population decreases cardiovascular risk.
I don't know why that would be.
Just like I don't know why frauds who call themselves "libertarians" voted for a very dangerous Statist socialist like Obama.
Do tell...
Both the increase in obesity and the increase in life span are the result of global warming. It's going to be catastrophic--Just ask Al Gore.
One of the interesting things about physical exercise is that it improves health and decreases mortality even if it *doesn't* produce weight loss. That is, a rise in rates of physical exercise will lengthen life even if it doesn't change rates of obesity at all.
It probably *is* true that being overweight or obese worsens health (slightly) even independent of all the associated diseases that a person might get like hypertension, diabetes, obstructive sleep apnea, etc., etc., but it's been remarkably hard to prove that.