Clinical Briefs

With Comments from Russell H. Greenfield, MD, Dr. Greenfield is Clinical Assistant Professor, School of Medicine, University of North Carolina, Chapel Hill, NC; and Visiting Assistant Professor, University of Arizona, College of Medicine, Tucson, AZ.

Selenium Too Sweet? Supplementation and Type 2 DM

Source: Stranges S, et al: Effects of long-term selenium supplementation on the incidence of type 2 diabetes. Ann Intern Med. 2007;147:217-223.

Goal: To assess the effect of long-term selenium supplementation on the incidence of new onset type 2 diabetes mellitus (DM).

Study Design: Secondary analysis of a randomized, double-blind, placebo-controlled trial (part of the Nutritional Prevention of Cancer, or NPC, trial) during its blinded phase.

Subjects: People with a history of nonmelanoma skin cancer in the year prior to randomization, but without DM at baseline, seen in dermatology clinics in locales where selenium consumption is low (average daily dietary intake = 90 m; analysis on n = 1,202 with a valid selenium value obtained within 4 days of date of randomization, n = 600 receiving selenium).

Methods: Subjects were randomized to receive either a placebo tablet containing yeast or 200 micrograms selenium from high-selenium baker's yeast tablets. Baseline evaluation included collection of demographic information, anthropometric data and behavioral characteristics. Biannual evaluations were subsequently held during which changes in clinical status were documented and blood samples were obtained. Individual medical records were reviewed periodically. An initial report of DM came from 3 possible sources: self-report during the clinical interview, reported use of drugs for DM, and medical records. For analytical purposes, plasma selenium levels were divided at tertiles and at the median level.

Results: Over an average follow-up of 7.7 years, a total of 97 new cases of Type 2 DM were diagnosed, with a higher cumulative incidence in the selenium group (n = 58, hazard ratio = 1.55). The risk for type 2 DM was consistently higher across all subgroups for those taking selenium, except in the highest tertile of body mass index, where no difference between groups was identified. A significantly increased risk for Type 2 DM was found for those subjects whose baseline selenium levels were higher than the median value (hazard ratio = 2.50). An exposure-response gradient was found across all tertiles of baseline plasma selenium level, with a significantly increased risk for Type 2 DM in the top tertile (hazard ratio = 2.70). When results were stratified according to behavioral and anthropometric covariates, lack of benefit with selenium supplementation persisted.

Conclusion: Long-term selenium supplementation at a dose of 200 m/d does not prevent Type 2 DM, but may instead adversely affect glucose metabolism, and increase risk for the disease.

Study strengths: Sample size; no participants lost to follow-up; selenium content of each batch of pills was determined in laboratories; review of diagnostic documentation.

Study weaknesses: Diabetes was a secondary outcome of the original trial, and conclusions were based on exploratory analyses; self-reported diagnosis of DM; lack of generalizability (most participants were older Caucasian men — non-whites were excluded); as the authors note, lack of detailed information on unmeasured risk factors (eg, family history).

Of note: The goal of the NPC trial was primarily to determine whether selenium supplementation offered benefits with respect to cancer prevention; insulin resistance, impaired glucose tolerance, and Type 2 DM are all linked to oxidative stress, and observational data suggest that dietary or plasma antioxidants protect against development of Type 2 DM; however, the few clinical trials of antioxidant supplementation to help prevent Type 2 DM or its complications have produced negative results; the data assessed in this paper are at least 16 years old; for inclusion, participants had to have a life expectancy of least five years and no history of internal cancer over the prior five years; non-white persons were excluded because the trial focused on the effects of selenium on nonmelanoma skin cancer, theoretically controlling for the effects of skin pigmentation on cancer recurrence risk; the incidence of DM in this trial mirrors that found in other studies of mainly white subjects; some data suggest that selenium may help prevent non-skin cancers, but recent research does not support a significant role for selenium with respect to cardioprotection.

We knew that: The majority of multivitamin and mineral supplement products contain 30-200 m of selenium; findings from animal models suggest that low-dose selenium supplementation improves glucose metabolism, while the effects of high-dose selenium supplementation remain unclear; both in vivo and in vitro studies suggest that selenium mediates many insulin-like actions, thereby enhancing insulin sensitivity; some data suggest that people with DM are relatively deficient in selenium, and that selenium may help prevent vascular complications in people with DM; the SU.VI. MAX study showed no benefit on fasting blood glucose levels from antioxidant supplementation that included selenium (100 m) over 7.5 years of follow-up, however longitudinal analysis of the relationship between baseline plasma selenium antioxidants and fasting blood glucose revealed a statistically significant association.

Comments: Selenium's importance in human health continues to be explored and debated. It plays a role in maintaining immunity and proper thyroid function, and there have been high hopes for selenium supplementation as a cancer chemopreventive agent, and as an aid in other endocrinologic realms, most notably DM. As regards the latter, animal data provided reason for optimism, but other research contradicts positive experiments. The current study brings up concerns of selenium actually increasing the risk of DM.

Unless a person suffers from significant gastrointestinal disease or requires total parenteral nutrition, most people in the United States have adequate, though perhaps not optimal, selenium levels. Trials such as the Selenium and Vitamin E Cancer Prevention Trial (SELECT) will help clarify matters regarding selenium supplementation, but results are not expected until 2013.

The conclusions of this study raise real concerns, but methodological limitations severely limit generalizability. It is far too early in this story to start lobbying for the removal of selenium supplements from store shelves, and benefits may yet be identified. Until more is known, no action need be taken, but consideration could be given to lessening supplemental selenium for those at high risk of developing DM.

What to do with this article: Keep a copy on your computer.

My Knee is Stuck! Acupuncture and OA

Source: Manheimer E, et al: Meta-analysis: acupuncture for osteoarthritis of the knee. Ann Intern Med. 2007;146:868-877.

Goal: To conduct a systematic review and meta-analysis assessing the efficacy of acupuncture in the management of knee osteoarthritis (OA).

Study design: Meta-analysis.

Study selection: Randomized, controlled trials (RCTs) longer than six weeks' duration that compared needle acupuncture with a sham, usual care, or waiting list control group for patients with knee OA of at least five years duration.

Methods: Data were culled from MEDLINE, EMBASE, and Cochrane databases without language restriction. In addition, unpublished data were obtained from eight authors. Outcomes were limited to pain and function (with an emphasis on WOMAC measures), and only those trials that employed insertion of needles into traditional meridian points were considered (additional insertion into tender points as well as electrical stimulation were deemed acceptable, but trigger-point therapy was not). In addition, trials that only compared two forms of acupuncture were not included in the analysis. RCTs were placed into categories according to control groups, which were sham, usual care, and waiting list. Acupuncturists assessing the data were blinded to results of the study as well as publication. Treatment adequacy was evaluated independently by the acupuncturists, who appraised four different aspects of acupuncture therapy: choice of acupuncture points, number of sessions, needling technique, and experience of the acupuncturist. Pooled effects of acupuncture were assessed to ascertain whether they met thresholds for minimal clinically important differences, defined as the smallest differences in scores that patients would perceive as beneficial.

Results: A total of 11 RCTs were included in the analysis, reflecting data on over 2,800 patients. No studies reported diagnosis according to the principles of traditional Chinese medicine, and all but one required radiographic confirmation of OA. Two RCTs used a flexible formula for point selection, eight used a set formula. The one remaining trial was pragmatic in nature, where both point selection and needling technique were left to practitioner discretion. Superficial needling alone was employed in one trial, while nine used deep needle stimulation. Electrical stimulation of the needles was used in four of the trials. Acupuncture was deemed adequate in all but the one pragmatic trial, where assessment was not possible due to individualization of therapy. Of the seven trials that employed a sham control, two used sham techniques that may have had physiologic effects. Four of the seven sham controlled trials were appraised to be of high internal validity, and together comprised the main source of evidence for the review (three of four also included a nonacupuncture group). With respect to efficacy, acupuncture provided clinically insignificant improvements in pain in the short-term and at six months when compared with a sham control. Acupuncture offered clinically relevant short-term improvements compared with patients who were in waiting list or usual care groups, improvements that were largely sustained at six months for the trials with usual care groups. As regards safety, only three RCTs described adverse events (AEs), and minor AEs occurred with equal frequency between acupuncture and control groups (no serious AEs were reported). Pooled effects of acupuncture were clinically relevant when compared with the waiting list and usual care controls, but not when compared with sham control.

Conclusion: Sham-controlled trials show clinically irrelevant short-term benefits of acupuncture for treating knee OA. Waiting list-control trials suggest clinically relevant benefits, some of which may be due to placebo or expectation effects.

Study strengths: Exclusion of trials of short duration; internal validity of studies assessed; use of minimal clinically important difference measures; examination of pooled effects of acupuncture; separate control group comparisons.

Study weaknesses: Problems inherent with meta-analyses (heterogeneity, etc.); challenge posed by potentially active sham interventions.

Of note: Exercise and weight loss are established beneficial non-pharmacologic interventions for OA, though some patients may have difficulty exercising due to their condition; the Western Ontario and McMaster Universities (WOMAC) Osteoarthritis Index pain and function measures are the most thoroughly validated instruments for assessing patients with knee OA; there is as yet no consensus regarding how best to assess treatment adequacy in RCTs of acupuncture; there are an estimated 400 active acupuncture points on the body, making physiologically inactive sham acupuncture challenging at the least.

We knew that: Data from large-scale RCTs suggest acupuncture to be an effective treatment for older people with knee OA, but meaningful conclusions remain elusive because the effects seem to differ depending on type of control group employed; OA is the leading cause of disability among older adults (with the knee being the most commonly affected joint); NSAIDs are only slightly better than placebo at providing short-term pain relief for people with OA, and are often associated with significant side effects (gastrointestinal bleeding is especially prevalent among elderly patients on NSAIDs); acetaminophen is even less effective in relieving OA pain than NSAIDs; acupuncture is safe in trained hands, with a low risk for serious side effects; a sham acupuncture control should be inert yet credible; PET scan research suggests that sham acupuncture may stimulate areas of the brain associated with natural opiate production.

Comments: This research team arguably comprises the North American leaders in the study of acupuncture, and while the article is not an easy read, its findings are fascinating. Many trials have floated the hypothesis that acupuncture therapy can reduce the pain of large joint OA, but results have been contradictory. Much of the contradiction may be traced to the type of control employed and, as the authors state, the impact of the meaning response or placebo effect, if not straightforward patient preference or pre-randomization expectation. Acupuncture appears to create a biological effect as evidenced by slight but measurable improvements in OA pain when compared with sham controls; however, some of the benefit of acupuncture for OA may stem from the placebo effect or emotional responses.

In considering positive research on acupuncture or CAM therapies in general, there are typically three camps: those who believe the intervention to be active and useful, those who believe the therapy to reflect "nothing more" than the placebo effect, and those withholding judgment until convincing evidence appears. What is intriguing is that the authors put forth that acupuncture is active and that patient expectation or the placebo effect are also at work. Indeed, as experts have promoted for any conventional medical intervention, to prudently engage the placebo effect may enhance therapeutic efficacy in association with biologic activity.

The authors note that acupuncture for the treatment of knee OA may be appropriate for short-term pain relief, but that it is too early to recommend that it be included as a routine part of treatment. This conclusion makes sense. The more notable aspect of the paper, however, is that it gently encourages us to consider additional possibilities regarding how any of our recommended therapies might work, even conventional ones proven biologically effective.

What to do with this article: Make copies to hand out to your peers.

Worth in Papworth? Breath work and asthma

Source: Holloway EA, West R. Integrated breathing and relaxation training (the Papworth Method) for adults with asthma in primary care: A randomized controlled trial. Thorax. 2007. [Epub ahead of print].

Goal: To evaluate the effectiveness of the Papworth Method (PM, a multi-component integrated breathing and relaxation technique) on people with asthma and dysfunctional breathing.

Study design: Randomized, controlled clinical trial.

Subjects: Adults (16-70 years) with mild-to-moderate asthma and no serious comorbidity who were recruited from a single primary care practice in semi-rural Great Britain (n = 85, 36 men).

Methods: All adults cared for by the medical group and with a prior diagnosis of asthma were contacted regarding participation in the trial. Interested patients who met study criteria continued to receive usual medical care, including routine asthma education, and were randomized to either a control group (n = 46) or to an intervention group who were to receive five Papworth treatment sessions from a respiratory therapist (n = 39). Assessments lasting approximately one hour each were made at baseline, and then six (approximately post-treatment) and 12 months later. The primary outcome measure was score on the St. George's Respiratory Symptoms Questionnaire (SGRQ); secondary outcomes included tallies of the Hospital Anxiety and Depression Scale (HADS), the Nijmegen dysfunctional breathing questionnaire, and objective measures of respiratory function (including spirometry). Relaxed breathing rate and end-tidal carbon dioxide were also measured.

Results: Post-treatment data were available for 78/85 subjects, and 12-month data for 72/85 subjects. Mean score on the SGRQ symptom subscale at six-months was 21.8 compared to 32.8 in the control group, with like findings at 12 months (24.9 for the active group, 33.5 for the control group). Total SGRQ scores, HADS and Nijmegen scores were all significantly lower in the intervention group. There were no significant between-group differences noted with respect to objective measures of respiratory function save for relaxed breathing rate. No adverse effects were reported.

Conclusion: The PM improves dysfunctional breathing, respiratory symptoms, anxiety, and quality of life when compared with usual care for adults with asthma.

Study strengths: Provides data on feasibility; multiple outcome measures.

Study weaknesses: Lack of generalizability (all subjects from a single, non-urban medical practice); small sample size; inadequate data imputation for those lost to follow-up; lack of information on changes in pharmacological treatment during the course of the trial as well as frequency of exacerbations; same person administered intervention and made assessments; added personal attention given to members of intervention group.

Of note: Dysfunctional breathing includes problems like hyperventilation and hyperinflation that are often seen in people with asthma; a Cochrane review on breathing methods for asthma concluded there was a trend towards clinical improvement when such methods were employed, but only seven small and disparate trials met inclusion criteria; usual care in this study did not include information on breathing exercises; the SGRQ is used to assess respiratory symptoms and associated quality of life issues, and provides a total score, as well as subscale information on experience of symptoms, their impact, and impairment in levels of activity; the Nijmegen questionnaire elicits information relating to hypocapnic symptoms like breathlessness associated with hyperventilation; logistical reasons were the most common cause for lack of participant follow-up; of the 612 adults with asthma in the practice, only 359 responded to a mailed survey regarding their condition and were then invited to a "physiotherapy-oriented" asthma assessment, of whom 142 responded, 85 of which comprised the study group; the PM technique was recommended for use at the first sign of asthma symptoms, as well as during the course of otherwise normal days.

We knew that: An estimated 300 million people worldwide have asthma; the cyclic wheezing and breathlessness seen with asthma is often complicated by anxiety; PM is best taught to patients during periods of remission; PM emphasizes five components — diaphragmatic and "nose-breathing," stress management, specific and general relaxation training, integration of breathing and relaxation techniques into daily living activities, and home exercises; a change of 3-3.5 points in SGRQ domains is considered clinically relevant.

Comments: As noted, a small number of trials have explored the possibility that mind / body therapies and specific breathing techniques might help ameliorate respiratory symptoms during asthma exacerbations. The underlying hypothesis is attractive — people understandably become anxious during an asthma exacerbation, often hyperventilate and then develop symptoms compatible with hypocapnia. If there were a way to help individuals calm themselves, perhaps even gain an added sense of control over the way asthma manifests, then emotional if not physical improvements might be experienced during acute episodes. The findings of this study support that hypothesis, but while the results are interesting, major study flaws make them far from definitive.

Breath work and stress management training might well be of benefit to most of our patients, if not ourselves. Provided asthma patients understand well when medical attention is required, training in select breathing and relaxation techniques would seem harmless and potentially helpful, but more out of commonsense than the results of existing research to date.

What to do with this article: Remember that you read the abstract.