People who regularly eat nuts have a lower mortality risk than people 
who don’t eat nuts, according to an industry-funded, observational 
study in the New England Journal of Medicine.

Researchers assessed the nut consumption of some 75,000 women in the 
Nurses’ Health Study and over 40,000 men in the Health Professionals 
Follow-up Study via regular food-frequency questionnaires. 
Participants were followed for up to 30 years.

People who ate nuts once a week had a lower risk for dying during 
follow-up than people who abstained from nuts (hazard ratio, 0.89). 
Risks decreased even more as nut consumption increased. In addition, 
more nut consumption translated to lower risk for dying from cancer 
and heart disease. The benefits were similar for tree nuts and peanuts.

In two studies of older people, statin use was not associated with 
dementia diagnoses.

Several cross-sectional and case-control studies have suggested that 
statins protect against development of dementia. However, prospective 
studies are better suited to examine this relation.

In one study, researchers prospectively followed more than 4000 
nondemented residents of a Utah county (age, ≥65). At 3-year follow-
up, 185 cases of dementia were diagnosed. After adjustment for 
potential confounders, statin use at baseline or at follow-up was not 
associated with reduced risk for incident dementia or Alzheimer disease.

In another study, researchers prospectively followed more than 2000 
nondemented people in a Seattle HMO (age, ≥65) for about 7 years; 312 
developed dementia during this interval. As in the Utah study, 
multivariate analyses failed to reveal any association between statin 
use and incident dementia or Alzheimer disease. Even patients with 
long durations of statin use and those who used high doses did not 
experience reduced risk for dementia.

COMMENT

The findings from these prospective studies cast some doubt on the 
idea that statins prevent dementia. Supporting these results are two 
large randomized trials (lasting 5 and 3 years) in which researchers 
observed no differences in cognitive function between statin and 
placebo groups (Journal Watch Jul 30 2002 and Journal Watch Jan 7 
2003). The possibility remains that long-term statin exposure, started 
early in life, confers some protection against cognitive decline.

Allan S. Brett, MD reviewing Zandi PP et al. Arch Gen Psychiatry 2005 
Feb. Li G et al. Neurology 2004 Nov 9.

The largest systematic review to date to address this issue suggests 
no adverse effects of statins on cognition.

In 2012, the FDA issued a safety announcement, noting that statin use 
might be associated with cognitive impairment such as memory loss or 
confusion. To evaluate this association, researchers conducted a 
systematic review and meta-analysis of 57 studies, including 19 
randomized controlled trials, in which assessments of cognition among 
statin users were reported; FDA postmarketing surveillance data also 
were analyzed.

Moderate-quality evidence suggested no excess risk for dementia or 
mild cognitive impairment, and low-quality evidence suggested no 
excess risk for Alzheimer disease. Analysis of postmarketing 
surveillance data revealed similar cognitive-related adverse-event 
reporting rates for statins (1.9 per million prescriptions) and two 
cardiovascular drugs not associated with cognitive impairment: 
losartan (1.6 per million prescriptions) and clopidogrel (1.9 per 
million prescriptions).

COMMENT

Although this systematic review was limited by the small number of 
randomized trials in which this outcome was evaluated, the data do not 
suggest excess risk for cognitive impairment with statin use. 
Nevertheless, a rare idiosyncratic effect of statins on cognition 
remains possible for individual patients, even if studies of large 
groups of people show no adverse effect on average. Indeed, in its 
safety announcement, the FDA acknowledged that it drew substantially 
from post-marketing case reports of adverse events.

Jamaluddin Moloo, MD, MPH reviewing Richardson K et al. Ann Intern Med 
2013 Nov 19.

New research provides stronger evidence that the composition of gut 
bacteria influences obesity.

Three new studies involving mice and humans provide new and stronger 
evidence that the composition of gut bacteria influences obesity.

In a U.S. study, fecal material from four human twin pairs (each set 
included one obese and one lean twin) was transplanted into 
genetically identical mice that were raised in a germ-free 
environment. The mice that received gut microbiota from obese humans 
became obese and developed obesity-related metabolism traits: 
dyslipidemia, insulin resistance, and inflammatory markers. In 
contrast, mice that received microbiota from lean humans stayed lean. 
Because mice routinely eat each other’s feces, researchers tried 
housing obese mice with lean mice — the obese mice gradually became 
lean and lost the obesity-related metabolic changes, and their gut 
flora changed to resemble that of the lean mice.

In a study from France, researchers compared genes from the gut 
microbiomes of 123 nonobese and 169 obese humans. The nonobese humans 
typically exhibited a different composition of gut bacteria, although 
the differences involved just a few species. Certain bacterial species 
were linked to insulin resistance, fatty liver, and inflammation. In 
another French study, investigators found the same “obesity-
related” bacteria to be present in their 49 obese participants, 
particularly those with obesity-related metabolism traits. When 
participants were placed on energy-restricted diets for 6 weeks, their 
microbiome changed toward that seen in lean individuals. Returning to 
their typical diets for 6 weeks moved the participants’ microbiomes 
back toward the initial composition.

COMMENT

Collectively, these three studies provide strong evidence that the 
composition of bacteria in our gut affects whether we are obese — and 
whether we develop obesity-related metabolism traits. They also 
indicate that what we eat can affect our microbiome. The studies do 
not prove that the gut microbiome will predict which people will 
become obese or develop obesity-related metabolism. And they surely 
don’t prove that directly altering the gut microbiome of obese people 
will improve their health. However, many investigators already are 
interested in pursuing those possibilities.

Ridaura VK et al. Science 2013 Sep 6. Le Chatelier E et al. Nature 2013 Aug 29. Cotillard A et al. Nature 2013 Aug 29.

The American Heart Association should consider depression to be a risk 
factor for adverse outcomes following heart attacks, according to a scientific statement by an AHA committee.

The statement, published in Circulation, included a review of 53 
studies assessing prognoses following ACS. The authors acknowledge 
heterogeneity among the studies, but they conclude that “the 
preponderance of evidence” points to depression as a risk factor for 
all-cause mortality, cardiac mortality, and a composite of all-cause 
mortality and nonfatal cardiac events after ACS.

They write: “There is not yet any strong evidence that treating 
depression improves survival after ACS; however, worsening depression 
is associated with worse clinical outcomes, and severe or persistent 
depression is reason enough to consider more comprehensive evaluation 
and treatment or referral to a mental health specialist.”

Circulation article (Free PDF)

Women with BRCA1 or 2 mutations who undergo oophorectomy have 
dramatically reduced risks for all-cause mortality and for cancers of 
the ovaries, fallopian tubes, or peritoneum, according to a Journal of 
Clinical Oncology study.

Researchers followed some 5800 women with a mutation in either gene 
for an average of 5.6 years. Relative to those with intact ovaries, 
those who self-reported having undergone either bilateral oophorectomy 
or salpingo-oophorectomy had an adjusted hazard ratio for ovarian, 
fallopian tube, or peritoneal cancer of 0.20. Their hazard ratio for 
all-cause mortality was 0.23.

Journal of Clinical Oncology

Jamaluddin Moloo, MD, MPH reviewing Salas-Salvadó J et al. Ann Intern 
Med 2014 Jan 7.
A Mediterranean diet supplemented with extra-virgin olive oil was 
associated with 40% reduction in risk for diabetes.

Weight loss through calorie-restricted diets and exercise lowers risk 
for type 2 diabetes. To assess whether Mediterranean diets without 
caloric restriction also protect against diabetes, researchers in 
Spain analyzed data on the 3500 nondiabetic participants in the 
PREDIMED prevention trial (NEJM JW Gen Med Mar 12 2013http://
www.jwatch.org/jw201303120000001), in which adults with ≥3 
cardiovascular risk factors were randomized to one of three diets: a 
Mediterranean diet supplemented with extra-virgin olive oil (EVOO; 3–
4 tablespoons daily), a Mediterranean diet supplemented with mixed 
nuts (1 ounce daily), or a control diet (consisting of advice to 
reduce intake of all fats). No participants were advised to restrict 
calories or to increase physical activity.

During median follow-up of 4 years, 273 participants developed 
diabetes. Incidence rates were 16, 19, and 24 cases per 1000 person-
years, respectively, in the EVOO group, the nut-supplement group, and 
the control group. After adjustment for potential confounders, the 
hazard ratios for diabetes were significantly lower for the EVOO group 
(0.60) and slightly but not significantly lower for the nut group 
(0.82) relative to controls. No significant changes occurred in 
weight, waist circumference, or physical activity levels across groups.

COMMENT

In this study, a Mediterranean diet supplemented with extra-virgin 
olive oil lowered diabetes incidence without associated weight loss or 
increased physical activity. The mechanism by which a Mediterranean 
diet might lower diabetes risk is unknown, but, as the authors note, 
such diets might alleviate inflammation, oxidative stress, and insulin 
resistance.

The suspected deleterious effect of supplementing men’s diets with 
selenium or vitamin E in an effort to prevent prostate cancer has 
apparently been confirmed. The findings appear in the Journal of the 
National Cancer Institute.

Researchers examined data from the SELECT trial, which examined 
whether supplemental selenium, vitamin E, or both could lower prostate 
risk. They found that men with low baseline selenium levels (as 
measured in toenail samples) did not benefit from selenium supplements 
alone or combined with vitamin E. And in fact, those with higher 
baseline selenium showed a significantly increased risk for high-grade 
prostate cancer with selenium supplementation.

For their part, vitamin E supplements alone had no effect in men with 
higher baseline selenium levels, but they increased risks for all 
grades of prostate cancer among those with lower baseline selenium.

The authors conclude that men over 55 should avoid both selenium and 
vitamin E supplements that exceed recommended dietary levels.

The so-called Mediterranean diet emphasizes olive oil, fruit, nuts, 
vegetables, and whole-grain cereals, as well as moderate intake of 
fish, poultry, and wine with meals. One of the most highly publicized 
studies of 2013 was a large trial from Spain in which 7500 adults (age 
range, 55–80) were randomized to one of three diets: a Mediterranean 
diet supplemented with extra-virgin olive oil (at least 4 tablespoons 
daily), a Mediterranean diet supplemented with walnuts, almonds, and 
hazelnuts (30 g daily), or a low-fat control diet. Although 
participants had no clinically evident cardiovascular (CV) disease, 
inclusion criteria were either diabetes or multiple nondiabetes CV 
risk factors.

During median follow-up of 5 years, rates of the primary outcome 
(myocardial infarction, stroke, or CV-related death) were 
significantly lower in both of the Mediterranean diet groups than in 
the control group. Most of the difference was attributable to lower 
stroke rates in the Mediterranean diet groups
  (NEJM JW Gen Med Mar 12 2013).

Because the low-fat control diet was not dramatically different from 
the Mediterranean diet, some observers believe that the supplemental 
olive oil and nuts were primarily responsible for this study’s 
outcome. Notably, about 50% of participants in each group were taking 
angiotensin-converting–enzyme inhibitors, 40% were taking statins, 
and 30% consumed at least one glass of wine daily. These background 
characteristics likely had some effect on the relatively low CV event 
rate in all groups.

Use of testosterone therapy, marketed to consumers for a condition
vaguely labeled as “low T,” is increasingly common; , in a
short-term randomized trial published in 2010, the incidence of
adverse cardiovascular events was higher in older men with
comorbidities who received testosterone therapy (NEJM JW Gen Med Jul
15 2010).  In the past year, two studies shed light on those
concerns.

In one retrospective study, researchers identified 8700 male veterans
with, or at high risk for heart disease whose blood testosterone
levels had been measured and were <300 ng/dL; 1200 of these men
received testosterone therapy, and 7500 did not. During a mean 27-
month follow-up, myocardial infarction, stroke, or death occurred in
26% of testosterone patients and in 20% of those not receiving
therapy. Higher risk for adverse events with testosterone therapy
occurred regardless of the presence or absence of heart disease.(NEJM JW

Gen Med Nov 19 2013).

Very little is known about risks associated with testosterone
therapy, but that fact doesn’t seem to be inhibiting its vigorous
promotion by pharmaceutical companies. For now, clinicians are
confronted regularly with the need to balance demands of men who seek
to improve their strength, energy, and appearance with theoretical
(and now, empirical) concerns about long-term risks of testosterone
therapy.

In another study, 140 men (age, 40–70) with erectile dysfunction and
low testosterone levels received optimized sildenafil (Viagra) therapy and then
were randomized to add daily transdermal testosterone (5–15 g) or
placebo for 14 weeks. Erectile function improved with sildenafil, but
add-on testosterone led to no further improvement (NEJM JW Gen Med Jan
2 2013).