4-1 SCREENING FOR PROSTATE CANCER
      �Few issues are as controversial.� Adequate evidence is absent.
      An approach that recommends men should be fully informed of risks and benefits of screening and then asked to make up their own minds is disingenuous when it is clearly difficult for specialists advisors to know the best approach. At present patients are offered an �informed� choice. But, the most informed observer can point only to uncertainty.
      �Tests have no meaning without clarity of purpose.� The main aim of PC screening is not to detect cancerous tissue, but to identify men who are asymptomatic and would otherwise die or be disadvantaged by untreated PC in the future, perhaps in 10 to 15 years.
      The weakness of the case for generalized screening rests on the poorly defined nature of the group identified for treatment. Current predictive values are poor. We still do not understand the natural history of PC well enough to distinguish those in whom disease is likely to progress from those whose pathology presents limited risk in terms of function and survival.
      There is . . . �no justification for screening programmes that expose men who might never be aware of the pathological changes within their prostates, to uncertainties about outcome and to certainties about the disagreeable nature of the treatment process.� �Exposing healthy people to treatments with specific hazards and uncertain benefits is unacceptable�
      �On the basis of the evidence, national programs of prostate screening are not justified.� Screening is a striking instance of therapeutic optimism.

4-2 COLON CANCER SCREENING GUIDELINES
      The rationale for offering several choices for screening, which vary widely in their efficacy and cost, is the hope of getting patients and physicians to consider an initial test. �We wanted to present options so physicians and patients can find something mutually agreeable, and increase the screening rate.� �You get the biggest bang for the buck from that initial screen.�

4-3 SCREENING MEN FOR PROSTATE AND COLORECTAL CANCER IN THE UNITED STATES.
      Among men in the USA, PC screening is more common than CRC screening. Men who choose to be screened should be made aware of the known mortality benefit of CRC screening, and the uncertain benefits of screening for prostate cancer.
      Men who insist on PSA screening should be informed of the greater benefit of colonoscopy.

4-4 MAMMOGRAPHIC SCREENING FOR BREAST CANCER.
      Eight trials have been published. In patients between ages 50 and 69, all reports of studies comparing screening with no screening showed protective effects of screening�a statistically significant 20 to 35 percent reduction in mortality from BC.
      The downside: False positive results necessitate further investigation. Nationally, an average of 11% of screening mammograms are read as abnormal. BC is subsequently found in about 3% of these women ( 0.3% of all mammograms). Thus, a woman has about a 10% chance of a false positive result with each mammogram. Because women are screened repeatedly, the risk of a false positive increases over time. One study estimated that, after 10 mammograms, about half of women age 40 to 64 will have had a false positive leading to needle biopsy or open biopsy in about 20%. The malpractice climate in the USA may work to increase the numbers of false positive reports.

4-5 PROSPECTIVE STUDY OF ALCOHOL CONSUMPTION AND RISK OF DEMENTIA IN OLDER ADULTS
      Compared with abstinence, consumption of 1 to 6 drinks weekly was associated with a lower risk of dementia among older adults.

4-6 WEIGHT LOSS COUNSELING REVISITED.
      If treatment success is defined exclusively as attaining ideal weight after losing a large amount of weight during a short term intervention, obesity treatment will almost certainly fail. �Obesity must be recognized as a chronic condition for which no cure can reasonably be expected.� However, even small weight loss can reduce obesity-associated risk factors for chronic diseases such as diabetes and hypertension.
      Physicians must remain non-judgmental and empathetic and distinguish between a �weight problem� and a �patient with a weight problem�. Recognize the challenges and frustrations the patient faces. Many obese patients, especially women, �have come to expect rejection and disparagement . . . and are favorably surprised when they receive the consideration that physicians usually accord patients�.
      An initial goal of weight loss of 10% of initial weight during a 6-month period is achievable and can significantly reduce obesity-related conditions. Most patients do not reach their ideal weight. Acceptance of a modest weight-loss is critical to prevent future disappointment. Emphasize the important health benefits and achievability of smaller reductions and help the patient accept a modest weight loss goal at the start of treatment.
      Obesity is a chronic condition requiring long-term care. For those not willing or able to lose, clinicians can help the patient to avoid further weight gain.
      Ultimately, prevention is the goal.

4-7 EFFICACY AND SAFETY OF LOW-CARBOHYDRATE DIETS
      Despite the abundance of lay literature on the topic of low-carbohydrate (Atkins�) diets, marked discordance exists between the knowledge needed to guide dietary choices and the information that is available in the medical literature.
      There is insufficient evidence to make recommendations for or against the use of low-carbohydrate diets. Among the published studies, participant weight loss while using low-carbohydrate diet was principally associated with decreased caloric intake and increased duration, but not with reduced carbohydrate content.

4-8 LOW-CARBOHYDRATE DIETS AND REALITIES OF WEIGHT LOSS
      �Without carbohydrate-containing foods (eg, breads) less fat is ingested because few people eat much fat by itself.� Thus, low-carbohydrate diets reduce calorie intake.
      A potential advantage of very low-carbohydrate diets is that removing sweets from the diet can reduce the gustatory stimulation that sweets produce. (This gustatory stimulation by sweets easily leads to over consumption.)
      The low-carbohydrate diets do reduce consumption of high fructose soft drinks, but do not deal with the preference that many humans have for sweets. �The aspect of carbohydrates as a preference in the diet is one that still needs to be addressed.�
      The broader issue of whether a unique diet exists that will produce long-term weight loss has yet to be evaluated. That �a calorie is a calorie� has been reaffirmed. The question of whether patients can adhere more easily to one type of diet or other remains to be answered.

4-9 ALDOSTERONE BLOCKADE AND HEART FAILURE
      �The addition of aldosterone antagonists to the regimens of patients with left ventricular systolic dysfunction and ongoing symptoms of heart failure despite optimal treatment with ACE inhibitors and beta-blockers can substantially reduce overall mortality and the rate of sudden death.�

4-10 A CLINICAL PREDICTION RULE TO IDENTIFY PATIENTS WITH ATRIAL FIBRILLATION AND A LOW RISK FOR STROKE WHILE TAKING ASPIRIN
      Stroke risk varies greatly in AF patients. This study sought to derive and validate a simple and easily applied clinical rule to identify individuals with non-valvular AF whose stroke risk while taking aspirin is low enough that oral anticoagulation is not necessary.
      Irrespective of age, a patient with non-valvular AF without previous stroke or TIA, without hypertension, without symptomatic coronary heart disease or heart failure, and without diabetes can take aspirin for stroke prevention and would not likely benefit from anticoagulation.
      Use of the rule would prevent almost one quarter of AF patients, regardless of age, to avoid anticoagulation. Sixteen percent of patients over age 75 were classified as low risk and thus would not be exposed to the risks of anticoagulation.

4- 11 PREVENTION OF CORONARY AND STROKE EVENTS WITH ATORVASTATIN IN HYPERTENSIVE PATIENTS WHO HAVE AVERAGE OR LOWER-THAN-AVERAGE CHOLESTEROL CONCENTRATIONS
      In absolute terms, benefits were small, with absolute differences between groups of 0.6% to 2.1%. (NNT [to benefit one patient over 3 years] = 47 to 166.)
      �Reaction to the 36% relative reduction in the primary endpoint and the other benefits observed in ASCOT may need to be tempered by consideration of the absolute risk reduction of a coronary event of 3.4 per 1000 patients-years.� �There are clearly financial implications.� (As well as adverse events from the drug. RTJ)

4-12 EFFECT OF MULTIVITAMIN AND MINERAL SUPPLEMENT ON INFECTION AND QUALITY OF LIFE
      A daily multivitamin-mineral supplement was associated with reduced incidence of patient-reported infection and related absenteeism in the subset of patients with type 2 diabetes who had a high prevalence of subclinical micronutrient deficiency.

1. The test is a false positive. No cancer is present.
2. The patient has a PC which is so low grade, so slow growing and likely would never lead to death or decreased quality of life.(Ie, the patient dies with and not from the disease.) Nevertheless, he will likely undergo radical treatment with a high likelihood of harm.
3. The patient has an aggressive, high grade, fast growing PC which remains confined to the prostate. Radical treatment may indeed cure.
4. The patient has PC, but it has extended beyond the prostate and is not curable. Medical therapy may benefit.

1. Persons at average risk should be offered screening for CRC and adenomas beginning at age50.
2. They should be offered options for screening, with information about advantages and disadvantages associated with each approach. They should be given an opportunity to apply their own preferences.
3. If a screen is positive, a complete colonoscopy should be advised.
4. The screening interval for double-contrast barium enema has been shortened to 5 years (vs the old recommendation of 10 years).
5. Low risk patients with removal of 1 or 2 tubular adenomas less than 1 cm in diameter are now advised to have a follow-up colonoscopy after 5 years rather than 3 years. (Adenomas with important pathological characteristics are unlikely to develop in 3 years.)
6. Patients with advanced or multiple (3 or more) adenomas should have a follow-up colonoscopy at 3 years.

1. In 2001 the Behavioral Risk Factor Surveillance System, an annual population-based telephone survey conducted by the CDC, gathered data on a representative sample of men above age 40.
2. Main outcome = proportion of men ever screened and up to date for both PC and CRC.

1. Men were more likely to report having ever been screened for PC than for CRC; 75% of those over age 50 had been screened with PSA vs 63% for any CRC screening test.
2. PSA screening was particularly common among elderly men; 69% of men age 70-79, and 56% of men age 80 and over¹ reported PSA screening in the past year.
3. Up-to-date PSA screening was also more common. For men age 50 to 69 (the group with the greatest acceptance of screening), 54% reported an up-to-date PSA, vs 45% with an up-to-date CRC screen.
4. Nationally, men in each age group were 15% to 20% more likely to be up-to-date on PSA than on CRC screening.

1. Although prostate and colorectal cancers are responsible for comparable numbers of deaths, CRC is responsible for a greater number of premature deaths. CRC accounts for 2 � times as many years of potential life lost before age 75 as PC.
2. Deaths from CRC were decreased by 14% to 33% in screened groups. �In contrast, no valid randomized trial evidence exists regarding the efficacy of prostate cancer screening.�
3. Why might men be more likely to have PSA screening? It is a simple blood test. Media messages have promoted PSA. Public personalities and even the US Postal Service (in the form of a stamp) have promoted PSA. Men may perceive themselves to be more vulnerable to PC. PC is more common (due in part to PSA testing). Men are more likely to know other men with PC than with CRC, and perceive themselves at higher risk.
4. It is likely that physician�s ordering practices are influenced by these factors, as well as perceived standards of care, meeting patient demand, and avoiding malpractice liability.
5. �Despite findings that most physicians are unsure about the benefits of PSA testing, the large majority routinely order the test.� �On the other hand, although most physicians are aware of strong evidence in favor of colorectal cancer screening, many refrain from recommending testing primarily because of anticipated patient resistance.�
6. �We recommend that physicians ensure that men who chose to be screened for cancer are aware of the known mortality benefit of colorectal cancer screening and the uncertainly about screening for prostate cancer.�
7. �Given the high rates of PSA screening among elderly men, physicians should ensure that elderly men (who are least likely to benefit and who may well be harmed by screening) make a well-informed decision about PSA screening.�

1. 1. �Mammography Service Screening And Mortality In Breast Cancer Patients: 20-Year Follow-Up Before And After Introduction Of Screening� Lancet April 26, 2003; 361: 1405-10 The 20-year study followed over 210 000 women. Conclusion��Mammography screening is contributing to substantial reductions in breast cancer mortality in two Swedish counties.�
2. 2. �Initiation Of Population-Based Mammography Screening In Dutch Municipalities And Effect On Breast-Cancer Mortality: A Systematic Review� Lancet April 26, 2003; 361: 1411-17

1. Cardiovascular Health Cognition Study entered over 3500 community-dwelling subjects in 1991-94. All were over age 65 and without cognitive impairment.
2. All received extensive cognitive testing at baseline. (Mean Mini-mental State Examination score = 28 of possible 30.)
3. Determined self-reported consumption of beer, wine, and liquor.
4. Followed to 1999 for incident dementia. (Mean follow-up = 6 years.)
5. Over the follow-up, 373 subjects developed incident dementia. (Mean age = 77)
6. The present study then compared alcohol consumption in these 373 cases with consumption by 373 matched controls who did not develop dementia.
7. Compared amount of alcohol consumption in the 2 groups.

1. Compared with abstinence, the adjusted odds for dementia in those consuming alcohol weekly:

   Alcohol intake less than one drink weekly	0.65
   1 to 6 drinks	0.46 (Lowest odds)
   7 to 13 drinks	0.69
   14 or more drinks	1.22

2. Of 373 subjects who developed dementia, 33 (8.8%) drank 1 to 6 drinks weekly. Of 373 subjects who did not develop dementia, 72 (19.0%) drank 1 to 6 drinks weekly. (By my calculation the absolute benefit in reducing risk of dementia associated with 1 to 6 drinks was 10 in 100 over 6 years. [NNT(to benefit one person over 6 years) = 10] )
3. There were generally similar relationships of alcohol use in Alzheimer disease and vascular dementia.
4. As expected, lower baseline MMSE scores, previous stroke, diabetes, and APOE e4 allele were more common in the dementia group.
5. The type of alcoholic beverage did not differ significantly.

1. �In this case-control study, moderate alcohol consumption had an inverse relationship with risk of dementia, even after multivariate adjustments and exclusion of former drinkers. Abstainers had odds of dementia that were about twice as high as the odds among consumers of between 1 and 6 drinks per week.�
2. Previous studies have reported moderate alcohol use is associated with lower prevalence of white matter lesions and subclinical infarcts (thought to be of vascular origin).

1) Anthropometric measurements:

A. Body Mass Index (BMI) weight in kg/ height in meters²
      Overweight --- 25 to 29
      Obese � 30 and above.

B. Waist circumference
      Men > 102 cm (40�)
      Women > 88 cm (35�)
(Abdominal fat accumulation is a marker for increased cardiovascular disease risk.)

2) Check for obesity related risk factors:
      Dyslipidemia, hypertension, smoking, impaired fasting glucose, physical inactivity,
3) Overweight patients and those with a large waist circumference who have 2 or more risk factors are candidates for weight loss.

Is the patient motivated to lose? Is he willing to invest the considerable time, effort and expense required?
Patients experiencing major life stressors, depression, or eating disorders usually need to address these issues first.

1. Given the heterogeneity of all trials, little can be concluded about the summary mean change in weight when all studies are combined.
2. When only randomized, controlled trials and the randomized cross-over trials are included, statistical analysis suggests that the studies are homogeneous. At the end of both lower- and higher-carbohydrate diets, participants� weight, body mass index, and percentage of body fat decreased.
3. Summary of mean changes in these randomized trials:

   	Carbohydrates in diet g/day
   	Less than 60 (7 trials)	Over 60 (75 trials)
   Weight changes over all	8 pounds	5 pounds
   Weight change
   Intake less than 1500 kcal/d	38	7
   Over 1500 kcal/d	11	3
   Diet duration
   15 days	13	3
   16- 60 days	12	8
   Over 60 days	5	2

4. Based on the data, it can be concluded that lowest-carbohydrate diets did not result in significantly greater weight loss than lower-carbohydrate diets.
5. When the 22 diets with the greatest mean weight loss (mean > 22 pounds) were considered, they varied widely with respect to carbohydrate content (range 10 to 271 g/day). However these diets 1) restricted caloric intake (mean = 1077 kcal/day), were of 2) longer duration, and included 3) participants who were significantly overweight at start (mean 222 pounds). �This suggests that these three variables may be more important predictors of weight loss than carbohydrate content.�
6. Overall, no change was found in any serum lipid levels. In the more homogeneous group of higher carbohydrate diets, there was a significant decline in total cholesterol levels, but not in other lipids.
7. No change was found in either fasting glucose or insulin levels among participants of either lower- or higher-carbohydrate diets�even among those with the greatest weight loss.
8. There was no change in systolic BP.

1. The data suggest that there is insufficient evidence to make recommendations for or against the use of these diets. Despite the large numbers of Americans who have apparently adopted this approach to weight loss, we know little of its effects or consequences. In particular, the diets have not been evaluated for use longer than 90 days, for individuals age 53 and over, or for patients with hyperlipidemia, hypertension, or diabetes.
2. The lowest carbohydrate diets (< 20 g/day; the recommended threshold for some of the most popular diets) have been studied in only 71 participants for whom no data on lipids, fasting glucose, and insulin levels, and BP were reported.
3. �We found insufficient evidence to conclude that lower-carbohydrate content is independently associated with greater weight loss compared with higher-carbohydrate content.�
4. Diets with restricted calorie intake and longer duration were associated with weight loss. �When lower-carbohydrate diet results in weight loss, it also is likely due to the restriction on calorie intake rather than carbohydrate intake.
5. Significant gaps in the published literature of low-carbohydrate diets exist: long term follow-up is inadequate; no data on effect of the diet among different races and ethnic groups; no data on degree of associated physical activity; no data on how adherence was measured and if counseling or other supportive measures were provided; no data presenting intention-to-treat (ie, many reported only those who completed the studies).

1) The low CBH diet does indeed lead to more weight loss during the first 6 months. Thereafter weight loss tapers off so the difference in weight loss is no longer significant at 12 months compared with the low fat diet.
2) There is a very high drop-out rate (~ 40%) even in subjects entering an enthusiastic trials. Attrition is high; adherence low. Intention-to-treat analysis is not possible. Bias is likely.
3) The weight loss may be modest compared with the degree of obesity in individual patients.
4) The effect on weight was due to a concomitant reduction in calorie intake.
(�Calories still count�) This could be due to a) satiety, b) monotony of the diet, c) ketosis. (One study reported that ketosis was prevalent during the first 6 months, but the authors did not consider it a significant factor in weight loss. (However, ketosis does blunt appetite. RTJ) 5) The low CBH diet was not associated with any deleterious effects on lipids, at least during the period of the studies. Insulin sensitivity improved.
6) Long-term safety and benefits are unknown.
7) �Any approach to caloric restriction that is not compatible with the daily lifestyle pattern is difficult to maintain long-term.�

1. Included over 2500 adult patients (mean age 70), all with non-valvular AF during 6 clinical trials.
2. All were free of stroke or TIA for at least 6 to 24 months.
3. All received aspirin (75 to 325 mg daily).
4. Primary outcome = stroke or TIA.

1. Overall, 166 patients (6.6%) had an event during over 4600 person-years of observation�equivalent to 3.5 events per 100 person-years. All were taking aspirin.
2. Determined characteristics of patients among the group associated with a low risk of stroke or TIA. Patients classified as low risk had all of 4 criteria:

   1) No previous stroke or TIA
   2) No treated hypertension or systolic BP > 139 mmHg.
   3) No symptomatic coronary heart disease
   4) No diabetes.
   

   (Other studies have found an increase in stroke for each of these risk factors in patients with AF.)
3. The prediction rule classified 24% of patients as low-risk.
4. Patients with low risk experienced 1 event per 100 person-years while taking aspirin. This compared with a rate of 1.2 events per100 person-years in the age and sex matched Framingham community cohort.
5. All other patients had an event rate of 4.2 per 100 person-year.

1. Using patient characteristics that are readily available at the bedside, the prediction rule identified AF patients whose annual rate of risk of stroke or TIA while receiving aspirin was comparable to that expected in the community. (Ie, a cohort presumably free of AF.)
2. Irrespective of age, a patient with non-valvular AF without previous stroke or TIA, without hypertension, without symptomatic coronary heart disease, and without diabetes can take aspirin for stroke prevention and would not likely benefit from anticoagulation.
3. Use of the rule would prevent almost one quarter of AF patients, regardless of age, to avoid anticoagulation. Sixteen percent of patients over age 75 were classified as low risk and thus would not be exposed to the risks of anticoagulation.
4. Since strokes are more common in patients with AF, and the benefits from aspirin in preventing stroke outweigh risk of gastrointestinal bleeding, it seems prudent to recommend aspirin for these low risk patients.

1. Multicountry, randomized study entered over 10 000 high risk patients (age 40 to 79; mean = 63) All had hypertension (defined as 160 systolic or more and diastolic as 100 or more; or both if untreated; or 140/90 or more if treated) and at least 3 other cardiovascular risk factors: left ventricular hypertrophy, ECG abnormalities, type 2 diabetes, peripheral arterial disease, previous stroke or TIA, male sex, age 55 or over, microalbuminuria, smoking, total cholesterol/HDL-c over 6. (This is a high risk group, except for only modestly increased cholesterols. RTJ)
2. All had a total cholesterol levels under 250 mg/dL. (Mean total-c = 210; mean LDL-c = 132 )
3. Randomized to:

   1) Atorvastatin 10 mg daily (dose not titrated) + antihypertension treatment, or
   2) Placebo + antihypertension treatment.

4. Primary end-point = non-fatal myocardial infarction + fatal CHD
5. Follow-up was planned for 5 years, but the study was stopped after a median of 3.3 years because of perceived benefit.

1. Atorvastatin lowered total cholesterol from a mean of 210 mg/dL to 161; and LDL-c from a mean of 132 to 89.
2. Changes in HDL-c were minimal.
3. BP was 138/80 in both groups at the end of the trial.
4.  

   Outcomes over 3.3 years	Atorvastatin	Placebo	Absolute difference	NNT 3 years
   (My calculations RTJ)	(n = 5168)	(n = 5137)
   Primary endpoint	100 (1.9%)	154 (3.0%)	1.1%	91
   (Non-fatal MI + fatal CHD)
   Total coronary events	178 (3.4%)	247 (4.8%)	1.2%	83
   Stroke	89 (1.7%)	121 (2.3%)	0.6%	166
   Total cardiovascular events	389 (7.5%)	486 (9.6%)	2.1%	47
   (All statistically significant.)
   Deaths	185 (3.5%)	212 (4.1%)	0.6%	166 ns

5. The benefit emerged in the first year. But no benefit was apparent among women (19% of cohort).
6. No significant adverse effects noted.

1. �In hypertensive patients who on average were at moderate risk of developing cardiovascular events, cholesterol lowering with atorvastatin 10 mg conferred a 36% reduction in fatal CHD and non fatal myocardial infarction compared with placebo.� 1
2. In absolute terms, benefits were small with absolute differences between groups of 0.6% to 2.1%. NNT [to benefit one patient over 3 years] = 47 to 166.
3. The apparent lack of benefit in women may have been due to chance. (The number of women entered was relatively small.)
4. �Current guidelines for the use of lipid-lowering agents vary strikingly around the world. Recommendations relating to treatment thresholds have been driven more by cost considerations than by evidence of treatment benefits.�
5. �Reaction to the 36% relative reduction in the primary endpoint and the other benefits observed in ASCOT may need to be tempered by consideration of the absolute risk reduction of a coronary event of 3.4 per 1000 patients-years.� �There are clearly financial implications.�

1. Randomized, double-blind, placebo-controlled trial, conducted in primary care clinics, followed 158 community-dwelling adults. All were considered to be relatively healthy.
2. Subjects were stratified by age 45 to 64 (n = 33) and 65 and over (n = 125); and by the presence of type 2 diabetes (n = 51)
3. Assessed nutritional status by a 3-day food diary. Defined nutrient deficiency as intake below the 33rd percentile of the recommended daily allowance for zinc, selenium, iron, folic acid, and vitamins A, C. E, or B6. (Choice of the 33rd percentile was arbitrary.) Thirty three subjects were considered deficient in one or more nutrients.
4. Randomized to: 1) multivitamin vitamin/mineral supplement; or 2) placebo taken once daily. The supplement was similar to that available over-the-counter (without a prescription) at drug stores.
5. Follow-up = 1 year.

1. More patients in the placebo group reported an infectious illness and absenteeism:
2. Overall

   	Placebo	Supplement
   Infectious illness reported	73%	43%
   Infection-related absenteeism	57%	21%

3. Subgroups

   Type 2 Diabetes patients ( n = 51)
   Infection reported	93%	17%
   Infection-related absenteeism	89%	0
   No diabetes (n = 79)
   Infection reported	60%	59%
   Infection-related absenteeism	35%	33%
   Age
   Under 65 (n = 33)
   Infection reported	78%	43%
   Infection-related absenteeism	58%	21%
   65 and over
   Infection reported	59%	44%
   Infection-related absenteeism	53%	19%

4. The infections reported were upper respiratory, influenza-like, gastrointestinal, lower respiratory, and a few urinary tract. Twenty percent experienced more than one infection during the year. None required hospitalization.
5. Quality-of-life scores did not differ between groups. Vitamin supplementation had no effect on physical or mental health scores.

1. Supplements had no effect in physical or mental health measures of quality-of-life. �Thus, it seems that multivitamin and mineral supplements do not enhance health or energy level anymore than does placebo.�
2. Benefit was observed almost entirely in participants with type 2 diabetes for whom the magnitude of benefit was dramatic.�
3. Diabetic subjects were more likely than non-diabetic subjects to be deficient in one or more micronutrients. However, the issue of how to define micronutrient deficiency remains unresolved. �We believe, however, that diabetic persons we studied had substantial preexisting micronutrient deficiencies.�
4. The lack of statistically significant benefit in those older than age 65 may have been because the number of subjects was small.
5. �Most participants correctly guessed what they were taking.� However, the investigators believed that this defect in blinding did not influence the results.
6. In certain diabetic samples, perhaps those with a high prevalence of micronutrient deficiency, daily use of a supplement can decrease infection frequency. �These dramatic results warrant testing in a larger clinical trial.