You don't need to be signed in to read BMJ Blogs, but you can register here to receive updates about other BMJ products and services via our site.

Archive for August, 2014

Primary Care Corner with Geoffrey Modest MD: Obesity paradox and stroke

27 Aug, 14 | by EBM

the “obesity paradox” found in several studies is that normal or underweight people have higher mortality, esp if they have chronic diseases. overall, studies with longer-term databases have found this to be less true, suggesting reverse causality: those with lower weight may disproportionately include people with underlying diseases causing them to lose weight. the current study, a large study with a remarkably good/complete database, adds to this (see doi:10.1001/jamaneurol.2014.1017). results:

–71617 patients admitted to Danish hospitals from 2003-2012 with a stroke, with information on BMI in 53,812. looked at deaths likely caused by the index stroke (if the death occurred within the first week or month after the stroke), and recorded in the Danish Stroke Register (coverage felt to be >80% complete).

–mean age 71.2, 47% women, 8% hemorrhagic stroke, mean BMI=25.7, with 10% underweight, 39% normal wt, 35% overwt, and 17% obese.

–within the first week of stroke, 4373 pts died with 3334 (4.7% of total) dying of stroke. no difference in risk of death from stroke by BMI (except in the underweight group)

–within first month, 7878 died and 5512 (7.7%) from stroke. no difference in risk of death from stroke by BMI (except in the underweight group)

–those with the lowest BMI tended to have the most severe stroke (stroke severity score), even controlling for age, sex, stroke subtype, cardiovasc risk factors, civil and socioeconomic status.

–BUT, there was an inverse relationship between BMI and age of admission for stroke — as compared to normal weight patients, those overweight had stroke 3 years earlier and obese 6 years earlier.

so, in general this study supports recommendation to achieve a normal body weight, despite the mixed messages from some of the other studies. the issue with underweight individuals continues to be muddied. this group (likely) includes normal underweight people (who might actually have a lower stroke mortality) along with some with significant medical or mental illnesses associated with both underweight and higher mortality.


Primary Care Corner with Geoffrey Modest MD: Hepatoma Screening in Chronic Hepatitis

26 Aug, 14 | by EBM

several medical societies suggest routine hepatocellular carcinoma (HCC) screening in patients with chronic hepatitis b or cirrhosis. a systematic review in annals of intl medicine, commissioned by the VA administration (see doi:10.7326/M14-0558), reviewed the data, finding “very low strength evidence” on the effects of screening on mortality. they found 22 adequate english-literature studies, with results:

 –1 large RCT done in China from 1993-95 included 9757 screened patients vs 9443 controls (screening with a-feto protein, AFP, and ultrasound every 6 months), found decreased HCC mortality of  83.2 vs 131.5/100K person-yrs, or rate ratio of 0.63 [95% CI 0.41-0.98]. but significant methodological flaws, including whether there was true randomization (concern that patients in each group did not have same HCC risk, questions about ascertainment of deaths in each group/selective reporting or analysis).

–another Chinese study from 1989-92 compared screening with AFP followed by ultrasound if AFP high (3712 pts) vs control (1869 pts). fewer screened patients had stage III HCC (19.8% vs 41.0%). HCC mortality similar in both groups (1138 vs 1114/100K person-yrs). BUT poor reporting of randomization/allocation, and only 28.8% of screening group completed all the prescribed tests.

–2 trials in Taiwan found no survival diff between shorter (3-4 month) vs longer (6-12 month) screening  (37.5% vs 6.7%), but HCC survival rates at 1-, 2-, and 4-yrs were not statistically different

–1 found no survival benefit to using AFP screening in those with hepatitis B

–18 observational studies from around the world found earlier-stage HCC with screening and generally higher survival from the time of HCC diagnosis, but unclear if reflected lead-time or length-time biases. for example, if they postulated a 90-120 day or longer tumor doubling-time , the survival advantage disappeared.

–harms of screening not assessed well in any of these studies, and other studies have found issues (eg a meta-analysis of 8 studies has found 2.7% with needle-track seeding from liver biopsy for suspected HCC. also reactions to contrast-enhnced CT or MRI. or other issues with biopsy, surgery, chemotherapies. or false positive results. or tumors that revert spontaneously. or….)

so… the data supporting aggressive screening (and medicalization, and cost) are not very good. a cochrane review for patients with hepatitis B in 2012 found insufficient evidence. another important issue in the US is that the RCTs (with their flaws) only dealt with patients with hepatitis B, and generalizing to other hepatitis (eg hep C, the more common one here) may not be appropriate, esp given the very different pathophysiologies of the different types of hepatitis. my sense is that, with the lack of conclusive evidence one way or another, we are stuck adhering to the guidelines about screening (which have largely dropped the AFP component, though there was a recent study suggesting some utility if the AFP were really high), with ultrasounds every 6 months or so. as always, would be better to have better data on which to make recommendations.


Primary Care Corner with Geoffrey Modest MD: NAFLD and food issues

20 Aug, 14 | by EBM

a couple of articles from the ny times.

One was on the increasing incidence of non-alcoholic fatty liver disease (NAFLD), which is increasingly becoming a significant cause of liver failure and transplantation, ironically as we move forward in eliminating hepatitis C as the major cause til now. the article notes that:

— NAFLD has increased dramatically of late (eg now in 10% of children and 20% of adults)

–2-3% of the US population have nonalcoholic steatohepatitis (NASH, the progressive form leading to cirrhosis)

–nationwide liver transplant increased from 1% due to NASH in 2001 to 10% in 2009 (and projected to be leading cause for transplants by 2020).

–NAFLD particularly common in Latinos, partly because they tend to have a genetic variant (PNPLA3), leading the liver to produce and store more triglycerides.  eg, liver dz is found in 50% of obese Latino children, and 25% of liver transplants in los angeles are because of NASH

–lots of comments in the article on lots of drugs-in-process, with relatively understated comments about lifestyle change, but see last year’s blog on effectiveness of wt loss and exercise. and, in this regard (as also per yesterday’s blog) the low glycemic index diet has particular value

and, speaking of food, last year’s article reviewed the rather distressing fact that the foods we currently grow have much less nutritional quality (fewer phytonutrients) than even in the recent past. wild dandelions of old had 7 times the phytonutrients than spinach (one of our better vegetables). purple potatoes from peru have 28x the anthocyanins (which may help fight cancer) than russet potatoes. why are the foods less beneficial? the author postulates that part of the issue is taste (some of the most beneficial phytonutrients are bitter, sour or astringent). and farmers over time tended to plant foods that have less fiber, more sugar/starch/oil, which were more palatale and provided fast calories for our strenuous lifestyle. eg, sweet corn is derived from a wild bushy plant with short spikes of grain instead of ears, with the kernals encased in dense shells, but had 10x the amount of protein of current corn. there have been natural genetic variations since then, but a major change took place over the past 50 years do to agricultural engineering, with supersweet corn (the predominant variety available in supermarkets) which are pretty devoid of nutritional value.



Primary Care Corner with Geoffrey Modest MD: Low glycemic index diets

19 Aug, 14 | by EBM

i will editorialize on an editorial in JAMA, which argues that adiposity may be the cause and not the consequence of overeating (see JAMA. 2014; 311(21): 2167). this editorial was written by david ludwig at boston childrens hosp, founder of the OWL clinic there (optimum weight for life), and longstanding champion of low-glycemic index diets (he wrote the seminal article on it — at least from what i’ve seen — in 2002, which reviews data and provides a theoretical framework  (see JAMA. 2002; 287(18): 2414). i suspect that the effectiveness data on low GI diets for weight loss, which is really no different from other diets, has more to do with the ability of patients to adhere to the diet long-term, as opposed to the specific diet itself. hard to adhere longterm when people are bombarded with calorie-dense, high carb foods (advertising, low-cost, high presence and availability in many areas including inner city and rural areas).  the gist of his argument is as follows:

–effectiveness of weight loss programs is okay short-term, but maintenance at 1 year is pretty miserable

–feeding studies show that changes in energy balance lead to compensatory mechanisms that antagonize the energy balance change (eg study of 41 people underfed or overfed to achieve 10-20% weight change led to compensatory change in energy expenditure — ie, underfed people tend to decrease energy expenditure as compensation to decreased energy intake and vice versa)

–the body has a pretty tightly controlled blood supply of energy (combo of major metabolic foods of glucose, nonesterified fatty acids, ketones), and any acute decrease or oxidation of them leads to intense hunger/food intake (an increase through fatty acid synthase inhibition or b3-agonist intake lowers food intake).

–insulin, an anabolic hormone, is important regulator of body weight: it drives glucose and non-esterified fatty acids into storage and is associated with weight gain. decreased insulin levels assoc with wt loss. fatty acid synthase expression is increased in fat tissue by insulin.

 –low-glycemic index diets decrease insulin levels (and, studies in rats given iso-caloric high vs low glycemic index diets develop hyperinsulinemia, and disproportionately more fat accumulation, even if put on food-restricted diet to prevent weight gain).

–there are also other factors which decrease anabolic drive, such as low refined-sugar intake, high polyunsat vs sat fats, low trans fats, high omega-3 intake, probiotics, etc; and physical activity, sleep and stress can affect calorie uptake and storage

–a small study by ludwig found that overweight or obese young adults achieving 10-15% weight loss with a variety of iso-caloric diets found that those on the low-fat (60% carbs) and low glycemic index (40% carbs) diets had decreases in both the resting and total energy expenditures as compared with the very low carb diet (10% carbs), and this decrease in energy expenditure tracked with the % of carbs in the diet.  ie, there seems to be a compensatory mechanism for weight loss by decreasing the energy expenditure, which is significantly less of a decrease with lower carb ingestion (see JAMA. 2012;307(24):2627-2634).

so, would be great to have real data showing the long-term effectiveness of low glycemic diet interventions. however, i find his arguments pretty compelling and have suggested low-glycemic index diets for many years now, to help with weight loss (mildly successful), diabetes control (pretty successful), and lipid improvement (pretty successful in decreasing triglycerides, increasing HDL, and improving cholesterol:HDL ratios). would be great if we could have public health programs which really promote healthy eating, increasing the accessibility/cost of healthy foods while decreasing those of calorie-dense junk and high carb foods…. geoff

Primary Care Corner with Geoffrey Modest MD: Heart failure interventions to decrease readmissions

18 Aug, 14 | by EBM

recent systematic review/meta-analysis of interventions to prevent readmissions for pts with heart failure (see Ann Intern Med. 2014;160:774-784). review was funded by Agency for Healthcare Research and Quality (AHRQ)


–heart failure (chf) is leading cause of hospitalizations (and health care utilization) in US

–25% of pts admitted with chf are readmitted within 30 days (35% of them readmitted with diagnosis of chf. rest with renal dz, pneumonia, etc)


–47 trials of diverse interventions to prevent readmission. pts with moderate to severe chf. mean age 70.

–at 30 days, high intensity home-visiting program (nurse or pharmacist usually) reduced all-cause readmission and combo of that plus death, though low quality of evidence

–over 3-6 months, home-visiting program and multidisciplinary heart failure clinics reduced all-cause readmission, with high quality of evidence

–structured telephone support (scheduled calls with structured questions, decision-support-software, etc) reduced chf-specific readmissions but not all-cause readmissions, with moderate quality of evidence

–home-visiting programs, multidisciplinary heart failure clinics, and structured telephone support reduced mortality

–BUT, neither telemonitoring (remote physiologic monitoring, eg with EKG, BP, weight, oxygen sats) nor educational programs (in person, or interactive computer-based) reduced readmission or mortality

so, not so surprising, though there is great local, regional, and national variability in post-hospital interventions. the results are certainly consistent with our anecdotal practice, that providing lots of nurse-based home visits (with occasional MD visits) and the availability of nurses to go to the house with minimal notice, decreases ER visits significantly and the seemingly inevitable readmission. seems that interventions with more human interaction (home visits, or even telephone support) do better than high-tech physiologic monitoring (telemonitoring). maybe not so surprising….


Primary Care Corner with Geoffrey Modest MD: Asthma and Early Exposure to Allergens

13 Aug, 14 | by EBM

lots of interesting articles over the past year on the microbiome and on allergen exposures and subsequent allergies. recent one of the Urban Environment and Childhood Asthma study (URECA), done in several urban areas (baltimore, boston, new york, st louis), a study of 560 kids at high risk of developing asthma and followed since birth for 3 years with environmental assessments including allergen exposure.  also, a second study looked at a nested case-controlled study of 104 kids assessing the bacterial content of their house dust exposure (see results:

–cumulative allergen exposure over first 3 years of life associated with allergic sensitization and related to recurrent wheezing

BUT, first-year exposure to cockroach, mouse and cat allergens was negatively associated with recurrent wheezing (ORs of 0.60, 0.65, and 0.75 respectively and all with p<=0.01). also, the probability of asthma decreased additively with increasing numbers of these allergen exposures. however, house dust mite and dog allergens had no association with asthma at 3 years old.

–differences in house dust bacterial content in the first year was associated with differences in atopic wheezing, esp with reduced exposure to Firmicutes and Bacteriodetes being associated with increased atopy. conversely, exposure to high levels of these allergens was associated with less asthma. the combination of high exposure to these bacteria in combination with high allergen exposure was additively protective of developing asthma. Presumably exposure to these bacteria protects against atopy since they come from bacterial families which are human colonizers and important producers of immunomodulatory metabolites, with presumption that early-life exposure to house dust containing these bacteria could innoculate the GI and/or respiratory microbiomes.

i think we are getting closer to the full story of the relationship between allergen exposure and allergies in kids. it has been hard to reconcile the many studies finding increased asthma in inner city environments (and the above study found that an annual family income <$15K was associated with increased asthma) with studies showing decreased asthma in those with high early allergen exposure. this study suggests that there may be important differences depending on the specifics of exposures, and that these exposures specifically need to be within the first year of life (presumably in the setting of an immature immune system). other microbiome studies suggest that the specific foods eaten can affect the specific bacterial content of the gut (and the lung, in some cases) and effect clinical issues such as airway resistance. of course, there are other precipitants for allergic asthma as well, including stress and exposure to other indoor pollutants.


Primary Care Corner with Geoffrey Modest MD: Azithromycin and Decreased Mortality

12 Aug, 14 | by EBM

there have been a couple of warning-type articles in the past couple of years about adverse effects of azithromycin, especially cardiac deaths. azithromycin is a really important antibiotic, but should be targeted only for significant bacterial infections — and it seems mostly to be used unnecessarily. a recent large cohort study did find some increased likelihood of MI but, most significantly, a lower 90-day mortality (see doi:10.1001/jama.2014.4304).  those on azithro were on combo therapy. others received guideline-concordant therapy. results:

–retrospective study from 2002-2012 of 73,690 patients older than 65 (mean age 77.8) from 118 VA hospitals who were hospitalized for community-acquired pneumonia, comparing 30- and 90-day outcomes, and matching 31863 on azithromycin vs same number not. high levels of comorbidities: 40% smokers, 52% with COPD, 35% with diabetes, 25% with prior malignancy. 15.9% were admitted to ICU

–no significant difference in identified potential confounders between groups

–90-day mortality 27% lower in those on azithro (17.4% on azithro, 22.3% not — ie, 4.9% absolute diff, with number needed to treat of 21)

–however, increased odds of MI (17% higher on azithro, 5.1% vs 4.4% — ie, number needed to harm of 144), though no diff in “any cardiac event”, which was 43.0% vs 42.7%, cardiac arrhythmias (25.8% vs 26.0%) or heart failure (26.3% vs 26.2%).

so, very large study, though retrospective. they did try to use propensity score matching to try to control for likely confounders (which, per their analysis, they did well), and as a large study gives lots of support for their conclusions. interesting that there was no increase in arrhythmias in the azithro group, surprising given the likely increase in QTc as suggested below (reminds me of the issue of prolonged QTc with citalopram, also with lack of anticipated clinical endpoints). bottom line: azithro is an important antibiotic to use, but only for significant probable bacterial infections. this also supports the american thoracic society (ATS) and infectious diseases society of america (ISDA) recommendation for patients with comorbidities (as in this study) to use either fluoroquinolone or combo macrolides (azithro is probably the best) with b-lactam (eg high dose amox, amox/clavulanate) to treat community-acquired pneumonia. of note, a couple of studies have found that combo fluoroquinolone/macrolide is better than fluoroquinolone alone, and that b-lactam plus macrolide is better than b-lactam plus fluoroquinolone.



Primary Care Corner with Geoffrey Modest MD: Meds for Alcohol Relapse

11 Aug, 14 | by EBM

a comprehensive systematic review and meta-analysis of studies looking at pharmacotherapy for alcohol use disorders (which are associated with 3x increase in early mortality, only 30% of patients receive treatment, and <10% are prescribed meds to decrease alcohol consumption) – see Reviewed 123 studies, almost all RCTs, with 22.8K participants. Mean age in 40’s. studies from 12-52 weeks long. Mostly enrolled patients after detox or at least 3 days off alcohol. Results (significant findings only):

–acamprosate tid (a glutamine antagonist and γ-aminobutyric acid agonist): NNT (# needed to treat) to prevent return to any drinking=12;

–naltrexone 50mg/d: NNT to prevent any drinking =20; NNT to prevent return to heavy drinking=12.

–In trials comparing these 2 drugs – no signif difference

–injectable naltrexone: no effect on return to drinking or heavy drinking, but 4.6% reduction in heavy drinking days

–nalmefene (one tab daily, off-label):  decrease of 2 heavy drinking days/month

–topiramate (off-label): 9% fewer heavy drinking days

–adverse effects: naltrexone and nalmefene – withdrawal from trials higher, esp for dizziness and gi effects  (# needed to harm=48). No signif adverse effects for acamprosate or topiramate

–they did look at studies of other off-label meds (aripiprazole, atomoxetine, desipramine, fluvoxamine, gabapentin, imipramine, olanzapine, ondansetron, and paroxetine – all with only one trial. Multiple trials for baclofen, buspirone, citalopram, fluoxetine, quetiapine, sertraline, valproic acid and varenicline), insufficient evidence to support their use.

–No significant benefit of disulfuram in the 4 studies found (data inconsistent. Some found that fewer drinking days in those returning to drinking. Also small numbers of patients).

–Nalmefene and topiramate considered  “somewhat effective”, but disulfiram is not.

–in many of these meds, improved results when combined with psychosocial intervention, esp studies with acamprosate and oral naltrexone.

So, pretty striking how little we are using meds, some of which have pretty good results for a very difficult, chronic, debilitating condition for many patients. Acamprosate and oral naltrexone are the best, with acomprosate being a t.i.d. med (less convenient) and contraindicated with severe renal impairment. Naltrexone being more convenient (once a day), but contraindicated if acute hepatitis, liver failure, concurrent opioid use, and with more adverse effects. The potential benefit of increasing treatment coverage to 40%, as per the authors, would reduce alcohol-attributable mortality by 13% in the European Union. Cost of a month supply of acamprosate or naltrexone is $140. ?cost of nalmefene (prescribed as one pill daily, but i am unable to get a price for it). best if meds supplemented with psychosocial intervention.


Primary Care Corner with Geoffrey Modest MD: H Pylori Eradication

8 Aug, 14 | by EBM

i posted in July (see here) about the likely utility of treating asymptomatic h pylori infections in high risk populations to decrease the risk of gastric cancer. there has been a subsequent systematic Cochrane review in the BMJ on this subject (see doi: 10.1136/bmj.g3174). 6 RCTs were identified, with inclusion criteria including using eradication therapy at least 7 days and minimum of 2 years of followup. primary outcome of occurrence of gastric cancer. results:

–all but one study conducted in East Asia (other in Colombia), so hard to assess effect in Western populations. only one study with longer-term followup (14.7 years). h pylori eradication rates in low 70% range|
–51 (1.6%) gastric cancers in those receiving eradication therapy vs 76 (2.4%) in 3203 control patients, for  RR 0.66. with assumption that eradication leads to lifelong benefit, the number needed to treat varies from 15 for Chinese men to 245 for US women (this is based on life-time risk of gastric cancer).
–of these gastric cancers, there was only one case of MALT lymphoma (not included in analysis)
–only 3 cases of esophageal cancer, so no significant difference
–only difference in adverse effects was rash in 3.1% on eradication therapy vs 0.1% on placebo

seems to me that the assumption of lifelong protection from h pylori eradication (ie, not get recurrent infection) is borne out by other studies. as noted in blog below, the benefit of eradicating the h pylori infection was in those without precancerous lesions at the time, which suggests that longer followup might yield even better gastric cancer prevention.  so, to me this reinforces the comment to test and treat asymptomatic patients who come from high prevalence h pylori countries (which also seems to track with those at higher risk of gastric cancer).



Primary Care Corner with Geoffrey Modest MD: Marijuana Adverse Effects Review

6 Aug, 14 | by EBM

there have been several concerning studies about the adverse effects of marijuana, of increasing importance now as legalization becomes more of an issue (which is not to say that other legal substances, such as alcohol or cigarettes, are better than marijuana. just that there are several newer studies showing that adverse effects of marijuana are more impressive than was understood before, and, with more widespread access, may lead to significant morbidity). in addition, the potency of marijuana has increased dramatically (THC content in 1980s of 3%, now it’s 12%). the natl institute on drug abuse (NIDA), a division of NIH, just published an overall review (see DOI: 10.1056/NEJMra1402309). Findings:

short-term effects:

–impaired short-term memory, making it difficult to learn and retain information
–impaired motor coordination, increasing risk of injury (esp motor vehicular). relationship between blood THC levels and performance in driving-simulation studies. more car accidents if use marijuana, but alcohol is worse.
–altered judgment, esp increase in high-risk sexual behaviors
–paranoia and psychosis, esp with higher doses

long-term effects:

–risk of addiction: approx 9% of users become addicted (per DSM-IV criteria), increasing to 1 in 6 in those who start as teenager and to 25-50% of those who smoke daily (2.7 million people in US over age 12 meet criteria for addiction, and 8.6 million meet addiction criteria for alcohol). withdrawal symptoms include irritability, sleeping difficulties, dysphoria, craving, anxiety. in general, adolescents seem to be more susceptible to adverse long-term effects, perhaps because the brain, including the endocannabinoid system, is actively developing in adolescence. cannabinoid dependence is 2-4 times more likely if begin marijuana as adolescent vs adult.

–brain development: brain is actively developing until around age 21. adults who smoked marijuana regularly as adolescents have fewer neural connections in specific brain regions, including the precuneus (a node involved in higher integration functions, such as alertness and self-conscious awareness), the fimbria (part of hippocampus important for learning and memory), prefrontal networks (for executive function), and subcortical networks (which process habits and routines). and, not surprisingly, frequent findings of decreased IQ in frequent users. also, poorer educational outcomes and increased likelihood of dropping out of school.

–gateway drug: unclear if marijuana use leads to other drug use, or if those inclined to drug use just start with marijuana (cheap, accessible). but, some of the animal studies are concerning: those exposed to marijuana when young have alterations in their mesolimbic dopamine reward systems, which can prime the brain for enhanced responses to other drugs

–relation to mental illness: some data that there is increase in psychosis, esp in those with genetic predisposition, though hard to establish causality. heavy use associated with earlier first psychotic episode than otherwise expected.

–cancer: lung cancer possibly associated, but nowhere near level of cigarettes. seems to be more chronic bronchitis (and inflammation of large airways, increased airway resistance, lung hyperinflation) in chronic marijuana smokers. some increase in respiratory infections and pneumonia. association with vascular disease (MI, stroke), with noted acute changes in vascular resistance and coronary microcirculation.

positive effects (ie, the rationale for medicinal marijuana):

–stimulate appetite, esp in those with AIDS (though there are some data that marijuana could exacerbate HIV-associated cognitive deficits)
–combating nausea/vomiting, esp in those on chemotherapy
–helping relieve severe pain and some types of spasticity
–decreasing intraocular pressure in those with glaucoma
–anti-inflammatory effects –role in IBD or rheumatoid arthritis (some data on cannabidol, which is devoid of psychoactive effects)
–MS — some efficacy for neuropathic pain, sleep disturbance, and spasticity

so, the brain effects are particularly concerning. also, the old mantra that marijuana is non-addicting seems to be wrong. at the health center, we are certainly seeing lots of patients, adolescents but also 70 year olds, smoking marijuana regularly. a few are able to decrease or stop just by engaging them in discussion (more so, in my experience, in adults than teens). others may benefit from psychosocial support, including cognitive behavioral therapy and motivational interviewing.


EBM blog homepage

Evidence-Based Medicine blog

Analysis and discussion of developments in Evidence-Based Medicine Visit site

Creative Comms logo

Latest from Evidence-Based Medicine

Latest from EBM