Richard Lehman’s journal blog, 19 April 2009

Richard Lehman

This week, Richard immerses himself in diabetes (“What a mess”), before covering an extensive range of subjects from the polypill to suicide, IQ, hyperhidrosis, and “irritable” bowels – all the while planning to compile “The Good Death Cookbook” – maybe…

JAMA  15 Apr 2009  Vol 301
Diabetes: what a mess. You can get some idea of it from this week’s JAMA, which is devoted to the diseases which cause sugary urine. The DIAD study lumped together 1123 patients with “type 2” diabetes – mostly a cyclical process of insulin resistance and progressive beta-cell failure. This confers a risk of coronary artery disease which varies from enormously increased in younger women to modestly increased in older men. The aim here was to discover and treat asymptomatic coronary artery disease and compare outcomes at a median of about 5 years. There was no significant difference. But the event rates were lower than predicted, which is something that always happens in cardiovascular trials these days, because so many patients are already on protective treatments. And the screening test was frankly rubbish: adenosine-stress radionuclide myocardial perfusion imaging, which had a positive predictive value of 12%.

Most dementia is vascular; most of it occurs in old people; and more of it occurs in old people with diabetes (hazard ratio 1.6). That’s all very straightforward, but how does treatment affect this outcome? Well, it probably cuts both ways, but the evidence isn’t all that good. This study of integrated health care records in California traces 16 667 patients with type 2 diabetes between 1980 and 2002 and finds that severe hypoglycaemia carries a risk of dementia proportionate to the number of episodes. Something to bear in mind if you believe in tight glycaemic control. On the other hand, a study which has just appeared on-line in Diabetologia (Xu WL et al) found that poor fasting blood sugar control (>7.8mmol/L) was associated with an increase in Alzheimer’s disease in their Swedish cohort of  1 475 elderly people with type 2 diabetes followed up for 9 years.

Let’s switch to youngsters who suddenly get type 1 diabetes. There is a whiff of optimism around them: a cure may even be within our grasp. This would have to arrest and reverse the auto-immune process which ends in total beta-cell destruction, and promising results were reported in 2007 using autologous nonmyeloablative haemopoietic stem cell transplantation. I’m not going to try and explain this to you: you can look it up in JAMA 2007;297:1599. Sceptics raised their eyebrows at reports of insulin independence in these 23 patients with newly diagnosed type 1 diabetes, aged between 13 and 31; but this report confirms that this has lasted for a mean of 31 months in 12 of the patients. Moreover, they show increasing levels of C-peptide, a marker for beta-cell function; so a cure may be on the cards for some of them. But against this, the procedure carries a risk of severe infection, late endocrine dysfunction and infertility.

So can diabetes be cured? Here’s a two-page whistle-stop guide to all the ways we might do it. Currently we can cure a few patients with type 1 using whole pancreas transplants, and a lot of patients with type 2 using bariatric surgery. So many, in fact, that some kind of rationing will need to be imposed, and another article argues that we should start with a BMI over 50.

The last year has seen the publication of three good long-term randomised trials – ACCORD, ADVANCE and VADT – proving that glycaemic control below HbA1c of 7 does not improve outcomes in type 2 diabetes. The only room for argument is about so-called “microvascular” outcomes – a curious rag-bag category which ranges from microalbuminuria to renal death and from background retinopathy to blindness. There was a 21% reduction in “nephropathy” from five years’ tight control in ACCORD, but I personally would not bargain a higher risk of hypoglycaemia for slightly less detectable protein in my pee when I’m 70-something, which is what this actually means. Read all about it in the editorial I wrote with Harlan Krumholz in this week’s BMJ. Then compare it with this piece about the three studies entitled “Glucose control in type 2 diabetes: still worthwhile and worth pursuing”. As Harlan headed his e-mail to me, “you gotta be kidding.”

NEJM  16 Apr 2009  Vol 360
The Intelligence Quotient (IQ) is about the most criticised measurement in the whole of science, but faute de mieux it’s the chosen one in this important study of fetal exposure to antiepileptic drugs. I’ve never tried to join MENSA or to test my own IQ, but I imagine that’s easy-peasy compared with doing it in three-year-olds. Anyway, that’s what the NEAD study did on an interim basis to 309 children born to mothers taking sodium valproate, phenytoin, carbamazepine or lamotrigine. The take-home message (if you believe it) is that valproate is the drug most likely to impair IQ and lamotrigine the least likely. More data from more IQ tests in three years’ time.

Among medics the best known Italian group after the Mafia is called GISSI – Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico. Give yourself a treat, say it out aloud. For some reason, they decided it was worth studying the effect of valsartan on the recurrence of atrial fibrillation after cardioversion. Ecco, valsartan has no effect on recurrent AF.

The journals all have a bit about diabetes this week, and in this case it’s the genetics of type 1A diabetes. That’s plain type 1 to you and me, if you believe in these distinctions – for a sceptical view, download and keep Declassifying Diabetes, an entertaining editorial by EAM Gale, Diabetologia 2006;49:1989. Genome-wide studies have uncovered lots of risk loci, none of them worth committing to memory unless you are a seriously demented gene gnome. Essentially things have only got worse since 1976 when James Neel, a leading geneticist of the time, titled a book chapter “Diabetes Mellitus: A Geneticist’s Nightmare.”

Lancet  18 Apr 2009  Vol 373
Gah! Having waded through the diabetic mire for weeks, I’m not sure I can face another study with too many interventions and a mass of surrogate end-points. The original Polypill, you may remember, was a single tablet containing several ingredients, to be taken by everybody over a certain age, to save bothering with coronary risk scores. Here, by contrast, are several different Polycaps to be taken by a variety of different single-risk groups with ages between 45 and 80. The drugs in combination behave more or less exactly as you would expect them to. Simvastatin lowers cholesterol, aspirin reduces urinary thromboxane B2, atenolol lowers the pulse rate and blood pressure, ramipril and bendroflumethiazide just lower BP. The point being? I can’t tell you, because I can’t work it out. Even when it gets some real end-points, I can’t see what this study is going to prove.

The Lancet’s contribution to diabetic enlightenment this week is the CALDIRET study in which German ophthalmologists discover that calcium dobesilate does not prevent diabetic macular oedema. So that’s that for calcium dobesilate. But the editorial about the trial on p.1316 ends with an Icelandic riddle which I want you to think about:
“We should distinguish between the prevention of retinopathy and the prevention of diabetic blindness. Diabetic blindness can be reduced or prevented without preventing retinopathy. Systematic screening for diabetic retinopathy and preventive laser treatment for those who develop macular oedema or proliferative retinopathy reduces the rate of blindness to about 0.5% in the diabetic population, irrespective of the prevalence of retinopathy.”
If I read this right, the 25% reduction in “microvascular endpoints” reported in tight control group of UKPDS, which consisted largely of diabetic retinopathy, has no meaning for the patient important outcome, which is visual loss. Blimey, I am losing the will to live.

So on to suicide. Here’s a longish seminar on the subject, full of interesting detail. International differences in suicide rates are very striking. Social factors, the availability of methods, and even the media coverage of prominent examples all seem to have a stronger influence than medical interventions. Nonetheless, it’s possible that primary care detection and treatment of depression may have a small role in preventing suicide.

I was hoping that mitral regurgitation might provide an oasis of mechanical simplicity in a somewhat gruelling week of medical reading: but not so. The valve itself looks deceptively simple but it gets distorted by whatever is happening to the left atrium and especially the left ventricle. You can look at lovely coloured whooshes of blood on Doppler echo but knowing when to intervene and how is anything but simple. It is particularly tricky to work out the chicken-and-egg situation of “functional” MR in heart failure, where the ventricular remodelling distorts the valve, the leakage of blood back into the atrium distorts the atrium, and atrial dilatation further opens the valve. This is not good news for the failing heart. But intervene at your peril.

BMJ  18 Apr 2009  Vol 338
Here’s the abbreviated version of a Swedish study of the effect of exercise in middle-aged men on their eventual longevity. Doing exercise between the ages of 50 and 60 has no immediate effect on mortality, but look again ten years later and the effect is as large as giving up smoking. Alas, for me, it may be too late because (a) I have only one year left before 60, (b) if I increase my exercise now I may actually increase my immediate mortality risk and (c) I’m too busy writing these things.

If you’re the kind of doctor who looks after burns and scalds, here’s a nice clear clinical review. It seems that there haven’t been any major advances in recent years, and it’s high time we got some new ways to deal with large area burns.

The educational pages of the BMJ win out over the research studies almost every week, and I certainly learnt something from this “Patient’s Journey” piece about hyperhidrosis. I was vaguely aware that there was a treatment called iontophoresis but no patient of mine has ever accessed it and I was quite unaware of how life-changing it can be. On the other hand, it does have to be given once a week and nobody quite knows how it works. There is a useful list of support groups.

Obstructive sleep apnoea in adults: do you miss it? Read this short piece to find out. OK, this is just a short plug for our new series Easily Missed. Suggestions for new topics are always welcome.

Arch Intern Med  13 Apr 2009  Vol 169
“Irritable Bowel Syndrome” is a dustbin diagnosis and I long for some rational means of explaining it to patients and some rational way to treat it. When serological testing for coeliac disease became available in the mid-1990s, I was sure that a trawl through our IBS patients would yield rich pickings, and that was the basis for my work partner Harold Hin’s landmark case finding study published in 1999. But we actually failed to find an increased prevalence in IBS. Others have done larger studies since then, and here is a systematic review. On the whole, coeliac disease is four times as prevalent in people with “irritable” bowels.

It sometimes seems to me that I am the only person who doesn’t know what unhealthy food is. I even thought of compiling a book called The Good Death Cookbook, junking the idea that you can stay healthy by avoiding certain types of food. Eat what you like, provided it includes lots of fresh fruit and vegetables, and just the amount of energy you need. Here is an attempt to review all the evidence supporting a causal link between dietary factors and coronary heart disease. I am delighted to see that there is insufficient evidence to exclude any kind of food, including saturated fat, meat or eggs; salt isn’t even mentioned. On the other hand, there is good evidence to yum up monounsaturated fat and eat like a Mediterranean.

Having eaten your Mediterranean food, you need to burn it off. This cluster randomised trial was conducted in Bilbao and Toledo – more Atlantic than Mediterranean bits of Spain – and assessed exercise prescribed in primary care. It provides some evidence that Spaniards who take too little exercise take a bit more when it is offered to them by their GPs.

Plant of the Week: Koelreuteria paniculata

I love trees with shrimp-pink new foliage in the spring. The one most often grown is Acer pseudoplatanus “Brilliantissimum”, which lives up to its name for a short season, though it is better to grow the related clone “Prinz Handjery” for leaves that don’t brown off so readily, and also for its frequent yellow flower spikes. Another slow-growing aristocrat is the little horse chestnut which sounds like a haematological malignancy, Aesculus erythroblastos. This produces wonderful tiers of pink fingers followed by typical horse chestnut flowers a week or two later. But perhaps the best all-year value is provided by this small tree or large shrub, the Golden Rain Tree.

I say small tree, because that is what it is likely to be in central England; in a favoured European spot, it might reach 20m, and the one in Sydney Botanical Garden grows well above that. At this point in the English spring, the tree is bursting with flame-like projections of flesh pink, which in time will become dark green pinnate leaves of uniquely complex form. Then in August, it will cover itself in panicles of golden yellow flower. As autumn sets in, these will produce bladder-like fruits and the leaves will turn bright yellow. What more can you ask? Scent, I suppose, and blue flowers, but our climate will never, alas, support the jacaranda tree.

Make do with this, if you have space. And try throwing a viticella clematis up it, to flower darkly amidst its golden rain.

  • Pingback: Valuable Internet Information » Richard Lehman’s journal blog, 19 April 2009()

  • Dear Dr. Lehman,

    Thank you for your efforts to clear up the picture on type 2 Diabetes and the concept of tight glycemic control. It is not so obvious or intuitive to conciliate the real world of evidence and clinical results, with the theory.

    As you also explained, in their Editorial comparing the trials ACCORD versus ADVANCE (1), Dluhy and Mc Mahon pointed to the fact that in ACCORD there were significantly more deaths and adverse effects like severe hypoglycaemia and weight gain than in ADVANCE.

    The use of insulin and glitazones was also much more intense, which helps to explain hypoglycaemias (usually accompanied by adrenergic reactions like tachycardia) and weight gain. Both drugs have been recently associated with more cardiovascular adverse effects (2, 3).

    I consider that the increased risk can be explained by an increase in the Pulse Mass Index (Resting Heart Rate by Body Mass Index divided by 1730), which I reported years ago to correlate strongly with the global cardiovascular risk calculated by the Framingham Risk Score (4).

    Drugs like metformin, in opposition to insulin and glitazones do not provoke weight gain or hypoglycaemia, nor increase the Pulse Mass Index, and is well known to reduce mortality and cardiovascular complications.

    The effects of used and studied drugs on the Pulse Mass Index should be carefully observed, as also the concept and consequences of tight glycemic control.

    Prof. Enrique Sánchez-Delgado, M.D.
    Internist-Clinical Pharmacologist
    Director of Medical Education
    Hospital Metropolitano Vivian Pellas
    Managua, Nicaragua


    1. Dluhy RG and McMahon GT. Intensive Glycemic Control in the ACCORD and ADVANCE Trials. N Engl J Med. 2008;358:2630-2633

    2. Mellbin LG et. al. The impact of glucose lowering treatment on long-term prognosis in patients with type 2 diabetes and myocardial infarction: a report from the DIGAMI 2 trial. European Heart Journal.2008;29(2):166-176

    3. Choy CK et. al. Type 2 diabetes mellitus and heart failure. Pharmacotherapy. 2008 Feb;28(2):170-92.

    4. Sanchez-Delgado E. and Liechti H. Lifetime risk of developing coronary heart disease. Lancet 1999;353:924-925

    Competing interests: None declared

  • Mikael Rabaeus MD

    Thank you for a refreshingly critical view on all the “science” that is published these days. I particularly liked and agreed with your comment on the polypill trial which has given rise to quite a hype here in Switzerland.
    Clearly we would all benefit and like to keep you around for long. So you might still consider exercising and eating.
    A previous study (Turning Back the Clock: Adopting a Healthy Lifestyle in Middle Age The American Journal of Medicine (2007) 120, 598-603) showed rather convincingly the benefit from changing your lifestyle.
    As for the diet, I totally agree with you that forbidding a lot of different food items is not going to bring you anywhere. You can eat whatever you like for as long as you ensure eating the correct things in majority. Which leads us back to the so-called mediterranean diet variations, which in my opinion is clearly the only ony to have shown convincingly that you can influence prognosis. In addition, after 25 years in clinical practice in Switzerland, I notice that the only diet that the patients stick to WITH PLEASURE once they have succeeded to make the change, is the regional variation of the Mediterranean diet. Swiss mediterranean diet is not the same as the Italian or the Swedish one.

  • Kathy Grant

    How about polycystic ovary syndrome for the easily missed series?