[ed. note: Michael Jorrin, who I call Doc Gumshoe, is a longtime medical writer (not a doctor) who writes for us about medicine and health a couple times a month. He has agreed to our trading and disclosure restrictions, but does not generally write directly about investment ideas. His ideas, thoughts and words are his own, and you can see all his past pieces here.]
As with tidings in general, some are glad and some are not so glad. Let’s start out with one that’s definitely on the glad side of the ledger.
To wit: the incidence of type 2 diabetes is declining. Let me say that again. New cases of type 2 diabetes, the kind that was formerly called “adult-onset diabetes,” but has recently been reported to a troubling extent in teen-agers and even children, are occurring less frequently.
Here are some data.
- In adults of ages from 18 to 44:
- Between 1980 and 1987, the incidence was under 2 cases per 1,000 population.
- Between 1987 and 2009, the incidence increased to 4.3 cases per 1,000.
- Between 2009 and 2014, the incidence decreased to 3.3 cases per 1,000.
- In adults of ages from 45 to 64:
- Between 1980 and 1999, the incidence increased from 5.2 to 8.9 cases per 1,000.
- Between 1999 and 2008, the incidence increased to 14.3 cases per 1,000.
- Between 2008 and 2014, the incidence decreased to 10.5 cases per 1,000.
- In adults of ages from 65 to 79:
- Between 1980 and 1991, the incidence increased from 6.9 to 8.9 cases per 1,000.
- Between 1991 and 2003, the incidence increased from 8.9 to 14.3 cases per 1,000.
- Between 2003 and 2014, the incidence decreased to 12.1 cases per 1,000.
As you can see, the largest decrease was in what we might call “middle-aged” adults, although younger adults experienced a similar percentage decrease, and in older adults, the decrease began a few years earlier. These encouraging decreases have to be regarded in the light of the large and problematic increases over the longer span from 1980 to the present, and, of course, it’s hard to predict whether the decreases will continue. When the first Doc Gumshoe blog about diabetes appeared, back in June of 2014, I cited the prediction that if present trends continued, about one-third of US adults would be diabetic by mid-century, and those proportions might be higher in groups such as Hispanics, African-Americans, and American Indian and Alaskan Natives, where the prevalence is already considerably higher than in the US population in general.
The numbers I quoted above, about the decrease in incidence, are from the CDC National Center for Health Statistics, and although they were no doubt available earlier if I had diligently sleuthed, I think they were reluctant to put them out too prominently, lest they be accused of making too much of what might turn out to be a blip.
I also pointed out in that piece that we’re far from the most affected country, In China, the incidence of diabetes has increased from less than 1% in 1980 to 12% at present. It’s also estimated that half of the entire Chinese population are prediabetic, and that by mid-century fully 40% of the population will be diabetic.
But to return to the excellent news about the decrease in the incidence here in our own country, why might this be? The CDC isn’t publicly venturing an opinion, but here’s some statistics that provide at least a tempting association.
The per capita consumption of soft drinks in the US has dropped from 53 gallons in 2000 to 43.8 gallons in 2014. In percentage terms, that’s similar to the decrease in new cases of type 2 diabetes in the past ten years or so. Americans have been switching from soft drinks to bottled water, consumption of which has been on a rapid increase. (Speaking for himself, Doc Gumshoe remains an advocate of tap water, which comes from our own excellent well!)
Why might this association be considered causal? Many experts in diabetes consider sugary soft drinks to be particularly harmful in bringing about diabetes. That’s because these drinks evade the kind of insulin response that normally takes place when we eat bona fide food. I discussed this in some detail in the diabetes blogs last year, but the quick version is that our insulin response is bi-phasic. The basal insulin secretion goes on all the time, keeping blood glucose at more or less normal levels and helping cells metabolize glucose for energy. The prandial insulin response starts fairly quickly, about 15 minutes after food enters the digestive system. What happens is that molecules called incretins signal the pancreatic beta cells to release insulin, so that by the time glucose from that meal starts to enter the blood-stream, insulin is ready to do its job. But glucose from soft drinks starts getting into the blood before the soft drink gets into the digestive system, so blood glucose goes to very high levels. This in turn triggers the release of abnormally high levels of insulin. And those high insulin levels lead to resistance at the insulin receptors, which is the first step on the way to type 2 diabetes.
In short, the big can of soda pop is a whole lot worse than the big slice of apple pie.
Of course, the decline in new diabetes cases doesn’t bring it back to the level it was before the whole country went on the junk food binge, and it certainly seems as though a lot of the world is on a similar binge. But maybe we’re headed in the right direction.
Next, some good news tempered with some bad news
First, let’s put this in the context of the overall good news: According to the Centers for Disease Control, the U.S. death rate for all causes declined 43% between 1969 and 2013, from about 1,279 per 100,000 people to about 730. This was due to large decreases in the deaths due to heart disease, stroke, and cancer.
However, one group of our fellow Americans did not share in this good news. And it was not, as some of us might expect, underprivileged minorities. African-Americans and Hispanics experienced improvements in life expectancy.
The group that did not see an improvement in life expectancy was white middle-aged Americans with less than a college education. Not only did their life expectancy not increase, it declined substantially. Life spans for most Americans and for residents of most developed countries have continued to improve at the rate of about 2% per year, as death rates declined. Starting in 1990, the number of deaths per 100,000 population in France, Germany, the UK, Canada, Australia, and Sweden declined from a range of 350 to 450 to a range of 100 to 300 by 2013. This decline in the death rate was also the case for Hispanics in the US, whose overall death rate was lower than that of US whites. But the death rate for middle-aged US whites actually increased by 134 per 100,000 population, to about 425 per 100,000 population.
This increase in the death rate and decline in the life expectancy of US whites was mostly due to higher mortality in that group of white Americans age 45 to 54 who had not attended college. During the years when the life expectancy of every other cohort in the US was increasing, theirs was shrinking at the rate of about one-half percent per year. This increase in the death rate did not apply to US whites older than 65 years, but was limited to whites in their middle years.
Among the counter-intuitive consequences of this trend is that as a group, American women in that demographic now have a life-expectancy about five years shorter than their mothers.
Angus Deaton, one of the authors of the study (Proc Natl Acad Sci 12/8/2015) reporting this deeply troubling trend, pointed out that the number of deaths in this cohort that would not have occurred, if the trend had been in line with the overall US trend and the trend in other developed nations, was about half a million, about 40 times the total death rate from the Ebola epidemic. Deaton, who wrote the paper with his wife Anne Case, is, by the way, the 2015 Nobel Prize winner in economics.
The causes of this anomalous trend are not easy to understand. The immediate causes, according to the statistical data reported by Case and Deaton, are poisonings, chronic liver disease, and suicide. “Poisonings” does not mean the usual triumvirate of arsenic, cyanide, and strychnine as in the Agatha Christie canon, but alcohol and recreational drugs, whether the illegal kind or diverted prescription drugs. The increase in deaths due to poisonings of this type has gone from two or three per 100,000 population in 1999 to more than 30 per 100,000 in 2014. Deaths from suicides and chronic liver disease have also increased sharply. And in a certain way, these immediate causes are linked. Alcohol is frequently an underlying cause of chronic liver disease, and some liver disease has been linked to abusing prescription opioids that are formulated with acetaminophen, which is a known cause of liver disease particularly when taken with alcohol. And alcoholism and drug abuse are often involved in the depressive pathology that leads some individuals to suicide.
The authors speculate that the underlying causes of alcoholism, drug abuse, and suicide are related to economic uncertainty in the US, particularly for individuals in this cohort, which, because of limited education, typically have a lower income, and because of their age, are less likely to participate in defined benefit pensions. Here are some passages from their paper:
“The epidemic of pain which the opioids were designed to treat is real enough, although the data here cannot establish whether the increase in opioid use or the increase in pain came first. Both increased rapidly after the mid-1990s. Pain prevalence might have been even higher without the drugs, although long-term opioid use may exacerbate pain for some, and consensus on the effectiveness and risks of long-term opioid use has been hampered by lack of research evidence. Pain is also a risk factor for suicide. Increased alcohol abuse and suicides are likely symptoms of the same underlying epidemic, and have increased alongside it, both temporally and spatially.
Although the epidemic of pain, suicide, and drug overdoses preceded the financial crisis, ties to economic insecurity are possible. After the productivity slowdown in the early 1970s, and with widening income inequality, many of the baby-boom generation are the first to find, in midlife, that they will not be better off than were their parents. Growth in real median earnings has been slow for this group, especially those with only a high school education. However, the productivity slowdown is common to many rich countries, some of which have seen even slower growth in median earnings than the United States, yet none have had the same mortality experience. The United States has moved primarily to defined-contribution pension plans with associated stock market risk, whereas, in Europe, defined-benefit pensions are still the norm. Future financial insecurity may weigh more heavily on US workers, if they perceive stock market risk harder to manage than earnings risk, or if they have contributed inadequately to defined-contribution plans.”
I find their reasoning persuasive but not totally convincing: if the current economic difficulties are the underlying cause of the “epidemic of pain, suicide, and drug overdoses,” why have African-Americans and white Hispanics not been similarly affected? A possibility that occurs to me is that there is a sort of self-perception among the non-college whites that they have sunk below the middle class. African-Americans and Hispanics in a similar position may think of themselves as on the way up, but whites have taken in the widely-shared view that “college is for everyone,” and if they didn’t go to college they are somehow unworthy. The mothers of the non-college American women mentioned above probably didn’t go to college either, but that didn’t make them feel bad about themselves.
A puzzling correlation and what it might mean
A curious paper was published in JAMA Internal Medicine earlier this year. (Jena BA et al, JAMA Intern Med 2015;175:237-244). The title more or less tells the story: “Mortality and treatment patterns among patients hospitalized with acute cardiovascular conditions during dates of national cardiology meetings.” This was a retrospective analysis of 30-day mortality among Medicare beneficiaries who presented at hospitals with heart attacks (acute myocardial infarctions, or AMIs), heart failure, or cardiac arrest from 2002 through 2011. They compared the outcomes in patients who presented during the dates of the two major national cardiology meetings, when the senior physicians would be likely to be absent, with the outcomes in patients who presented on the same days of the week when no cardiology meetings were being held.
The motive for conducting this analysis is fairly evident. The obvious assumption is that if you go to the hospital with an acute heart condition that requires immediate treatment, you want the best doctors to be available to take charge of your treatment, and if the best doctors are out of town someplace attending a cardiology meeting, that poses a definite risk that you won’t be getting the best treatment.
The results were the reverse of that obvious assumption. Patients hospitalized with those acute heart conditions when the cardiology meetings were being held actually had better outcomes in terms of mortality, and the differences were substantial for heart failure and cardiac arrest. Thirty-day mortality for heart failure patients presenting during those cardiology meetings was about 30% lower, and in cardiac arrest patients it was about 15% lower. Both of these differences were significant – P < 0.001 for heart failure and P = 0.01 for cardiac arrest. There was almost no difference in outcomes for AMI patients, although, interestingly, patients presenting with AMI during cardiology meetings were about 25% less likely to receive percutaneous coronary intervention (PCI). This is a procedure formerly known as angioplasty, in which a stent is placed in a coronary artery to permit blood flow in a place where the artery has been obstructed. The JAMA paper prompted an opinion piece in the New York Times on 22 November 2015 under the provocative title “Are Good Doctors Bad for Your Health?” The writer, Dr. Ezekiel J. Emanuel, speculated that when the top doctors were away at the cardiology meetings, the doctors who stepped up to care for these patients were probably more recent medical school graduates. As he put it, “One possible explanation is that while senior cardiologists are great researchers, the junior physicians – recently out of training – may actually be more adept clinically.”
The only problem with this explanation is that is assumes that the physicians attending the cardiology meetings are the “good” doctors, while those left behind to mind the store are the junior doctors. This is not necessarily the case. Indeed, the physicians presenting papers at the national meetings of the American College of Cardiology and the American Heart Association are the researchers and the lead authors of studies. But large numbers of the less eminent physicians also attend, while many very experienced clinicians do not attend those meetings, choosing instead to stay home.
A possible explanation is the simple urge, on the part of some of those top-ranking cardiologists, to do more, or to try something different. Another is the possibility that the lower-ranked hospital clinical staff defers to the senior doctors and doesn’t insist on some of the checklist procedures or adherence to guidelines that would routinely be followed.
The value of checklists is evident, ranging from the necessity of verifying again and again whether the patient being treated is indeed the one whose chart the physician is looking at. Again and again patients will be asked their names and date of birth. It doesn’t do for the surgeon to do a liver biopsy on John Smith when the patient who needs the liver biopsy is Joe Brown. Checklists routinely run to fifty items or more, and some doctors may feel impatience at being pestered with simple reminders.
The guidelines issue is considerably more complicated. Guidelines are valuable and serve a clear purpose, without doubt, especially in cases of uncommon or rare diseases, where the clinician responsible for deciding on a course of treatment is not on sure footing, or where there is new information that hasn’t yet trickled through the entire healthcare community. Also – and here is where the Doc Gumshoe antennae deploy – guidelines inevitably take cost-effectiveness into account, sometimes to a major degree. Most famously, in my view, were the hypertension guidelines that continued to recommend diuretics for all newly-diagnosed persons with hypertension, mostly because they are cheap. Guidelines may also short-circuit a physician’s clinical judgment. The prudent clinician is apt to hesitate to institute a treatment that is not recommended by the guidelines, nor yet omit a recommended treatment.
On the subject of guidelines, here’s the first sentence from a paper in the December issue of the American Journal of Medicine (Sab S. et al. Am J Med 2015;128:e7-e9):
“Evidence-based guidelines for patient management cannot account for all clinical circumstances. This challenge is demonstrated by the presentation of a patient with nonischemic cardiomyopathy complicated by a large left ventricular thrombus.”
The patient was treated according to 2013 guidelines by the American College of Cardiology/American Heart Association for the treatment of heart failure. This guideline does not consider therapeutic anticoagulation except in cases of atrial fibrillation, nor does it consider surgical intervention to correct the thrombus. The guidelines-based treatment was not a success. At the conclusion of treatment, all arterial circulation below the right femoral artery had become blocked, and it was necessary to amputate his right leg above the knee.
This is not to say that the ACC/AHA guidelines were responsible for this patient’s unfortunate outcome. However, the fact that this group of physicians from the University of California at Davis’s Division of Cardiovascular Medicine chose to publish this paper strongly suggests that adherence to guidelines may not always lead to optimum outcomes.
So this brings us full circle to the puzzling data about the improvement in patient outcomes when the presumably best doctors were out of town. The suspicion, lurking just beneath the surface, is that those senior physicians, eminent researchers and thought leaders, are too smart and too arrogant to go by the book, but push new and untested therapies of those unlucky patients who fall into their clutches.
Or could it be that the attendees at these national cardiology meetings are not necessarily the best doctors? There is a fundamental difference between medical research and medical practice. The best researchers, and many of the people who follow that research, may be more scientists than they are doctors. Their attention may be more fixed on the data than on their patients.
I take issue with that headline in the Times, “Are Good Doctors Bad for Your Health.” Good doctors, doctors who are focused on the patient, are good for your health. Attention to the patient, and how the patient is responding to treatment, is the key characteristic of the good doctor. I don’t want the doctor to be looking at the computer, focusing on guidelines and algorithms. I want him or her to be looking at me.
And to close out the old year, a good reason to lift a glass in celebration
Just a couple of weeks ago, a paper was published in the British Medical Journal Open that concluded that moderate alcohol consumption was associated with lower mortality in patients with mild Alzheimer’s disease. (Berntsen S et al, BMJ Open:2015;5(12):e007851).
The paper was a post-hoc analysis of the Danish Alzheimer’s Intervention Study (DAISY), which was looking at the efficacy of psychosocial counseling and support in patients with mild AD. There were 321 subjects in the study, and the time period covered by the study was 36 months. The amount of alcohol consumed by the subjects was reported by their caregivers, and grouped into four categories: abstainers, occasional drinkers (one unit of alcohol per day or less), moderate drinkers (two to three units per day), and persons who consumed more than three units per day.
A “unit” of alcohol was considered to be 15 mL, or about half an ounce, of pure alcohol. One glass of wine, usually about five ounces, or 150 mL, would contain about that much pure alcohol if the wine had the typical alcohol content, around 10% – 12%. Younger, coarser wines may have up to 14% alcohol, but by the time they have aged in casks for a few years, the alcohol content has declined a bit. So, figure a unit of alcohol is about equivalent to a glass of wine. A shot of straight hooch is probably a bit more.
The results of this analysis were quite interesting, to say the least. The moderate drinkers, who drank the equivalent of two to three glasses of wine per day, experienced 77% lower mortality than the abstainers or the subjects who drank one unit per day or less. The results followed the “U-curve,” which I have mentioned in other blogs – that is, only the moderate drinkers derived any benefit. The heavier drinkers, who consumed more than three drinks per day, had about the same mortality rate as the abstainers and the occasional drinkers.
This was not a large study, nor did it extend over a long enough period to be genuinely conclusive. Of the 321 subjects, 53 died during the three-year period, and the majority of those 321 subjects were the occasional drinkers, so the basis is relatively small. But the size of the effect was considerable.
The results from the DAISY trial echo and reinforce a large analysis of 143 papers which examined the effects of alcohol on cognition. It was published in Neuropsychiatric Diseases and Treatment back in 2011. (Neafsey EJ, Collins MA, Neuropsych Dis Treat 2011;17:465-484). The papers reviewed were not singly focused on alcohol and cognition, but mostly evaluated alcohol consumption along with other factors. Of the papers, about half provided ratios between drinkers and non drinkers, and half rated cognition in drinkers to be rated as “better,” no better,” or “worse” than cognition in non-drinkers.
Once again, the results of this analysis were highly interesting. The moderate drinkers were 23% less likely to develop AD or other cognitive deficits as compared with non-drinkers. The heavy drinkers were likely to have more memory impairment and a higher risk for dementia than moderate drinkers; however, the data on this group was considered limited.
The researchers noted that the benefits of moderate drinking were primarily seen in the older subjects, those more than 55 years of age. In younger subjects no differences in cognition were seen between drinkers and non drinkers.
As with many meta-analyses, some of the conclusions are necessarily tentative, because the studies have a range of parameters that do not match up. Thus, while overall, wine was found to be more effective than beer or distilled spirits in slowing cognitive decline and the onset of AD, some studies that made the distinction between types of alcohol found no difference. The studies included came from 19 countries, and similar general benefits from moderate alcohol consumption were reported in 14 of those 19 nations, and in the remaining five, three reported benefits that did not attain significance.
The benefits of moderate alcohol consumption (I’m getting tired of putting the word “moderate” in front of “alcohol consumption” every single time, but I’m scared that if I don’t, I’ll be held single-handedly responsible for sending the entire population of Gumshoeland over the brink) have been known for about 20 years, since the BMJ published a paper back in 1996 reporting about a 30% reduction in overall mortality in moderate drinkers. I wrote about it in a Doc Gumshoe piece about the Resveratrol Bust, which you can find here. In that piece, I quoted what the Harvard School of Public Health had to say about the benefits of moderate alcohol consumption:
Possible Health Benefits of Alcohol
What are some of the possible health benefits associated with moderate alcohol consumption?
More than 100 prospective studies show an inverse association between moderate drinking and risk of heart attack, ischemic (clot-caused) stroke, peripheral vascular disease, sudden cardiac death, and death from all cardiovascular causes. The effect is fairly consistent, corresponding to a 25 percent to 40 percent reduction in risk.
Beyond the Heart…
The idea that moderate drinking protects against cardiovascular disease makes sense biologically and scientifically. Moderate amounts of alcohol raise levels of high-density lipoprotein (HDL, or “good” cholesterol), and higher HDL levels are associated with greater protection against heart disease. Moderate alcohol consumption has also been linked with beneficial changes ranging from better sensitivity to insulin to improvements in factors that influence blood clotting, such as tissue type plasminogen activator, fibrinogen, clotting factor VII, and von Willebrand factor. Such changes would tend to prevent the formation of small blood clots that can block arteries in the heart, neck, and brain, the ultimate cause of many heart attacks and the most common kind of stroke.
So, the benefits of alcohol with regard to cognition and AD shouldn’t come as a surprise. Of course, the focus of medical professionals would tend to be on mechanisms of action that can be assessed scientifically. But there’s another side to it. I strongly believe that a couple of glasses of wine with our dinner, or perhaps a few sips of a nice single malt, greatly enhance our pleasure and contribute to our joy of life – which, I’m convinced, might also help in keeping us alive and kicking.
Best to all, and enjoy every kind of holiday that comes around! Thanks, Michael Jorrin, (aka Doc Gumshoe).