Dissecting the AIM-HIGH Trial
One of the studies mentioned in The New York Times blog post was the Atherothrombosis Intervention in Metabolic Syndrome with Low HDL/High Triglycerides: Impact on Global Health Outcomes trial—called AIM-HIGH for short. Although The New York Times called this a new study, the results were first reported and published in 2011.1
However, AIM-HIGH returned to the spotlight when the authors recently wrote a letter to the editor in The New England Journal of Medicine in which they summarized the results of the study and compared the adverse events to those experienced by subjects in a newer trial that investigated the use of niacin combined with a drug used to eliminate flushing of the skin, a common niacin side effect.2
In the AIM-HIGH trial, researchers randomly divided a total of 3,414 subjects to receive 1,500 to 2,000 mg of extended-release niacin (1,718 subjects) or placebo (1,696 subjects). All patients in the study also received the statin drug simvastatin as well as the cholesterol-lowering drug ezetimibe if needed to keep cholesterol levels at 40 to 80 mg per deciliter (1.03 to 2.07 mmol per liter). The goal of the study was to determine if extended-release niacin would decrease the risk of cardiovascular events in patients who had established atherosclerotic cardiovascular disease and lipid problems (low levels of HDL cholesterol, high triglycerides, and small, dense particles of LDL cholesterol).
The study found that niacin in addition to a statin produced no clinical benefit—in other words it did not decrease the risk of cardiovascular events. This result was what earned the study the most attention. However, some researchers have proposed that the AIM-HIGH trial "may have been too small to detect plausible reductions in vascular events."3
After taking niacin for two years, the study subjects did experience some benefits. Niacin raised the median HDL "good" cholesterol level from 35 mg per deciliter (0.91 mmol per liter) to 42 mg per deciliter (1.08 mmol per liter)—a significant change. Furthermore, after taking niacin, the subjects' triglyceride levels fell from 164 mg per deciliter (1.85 mmol per liter) to 122 mg per deciliter (1.38 mmol per liter). Additionally, the LDL cholesterol level also fell from 74 mg per deciliter (1.91 mmol per liter) to 62 mg per deciliter (1.60 mmol per liter).
Another plausible reason why the study showed no difference in cardiovascular risk reduction may be because cholesterol is only one risk factor for cardiovascular disease. We also need to factor in C-reactive protein (CRP), homocysteine, fibrinogen, and the oxidation of LDL. Niacin is primarily used to lower lipids. Unlike statins, it has not been shown to lower C-reactive protein. Nor does it address other factors associated with cardiovascular risk reduction, which is why nutritionally minded doctors advise patients to address those other mechanisms of cardiovascular disease as well as metabolic syndrome, another risk factor for cardiovascular disease. In other words, niacin is part of a bigger picture that includes lifestyle measures and other supplements that can address CRP, LDL oxidation, homocysteine, and fibrinogen.
There were some adverse events in the AIM-HIGH study. These included the well-known side effects of niacin experienced by some people including itching, flushing, diarrhea, rash, thrombocytopenia (low blood platelets), and increased blood glucose levels as well as more rare adverse events including abnormal results on liver-function tests. Niacin also significantly increased the risk of serious infection in 139 subjects in the niacin group and 98 in the placebo group.
There was a low rate of serious hemorrhagic adverse events. The rate wasn't significantly different between the group taking extended-release niacin and the group taking the placebo (3.4 vs. 2.9 percent).
It is of interest to note that myopathy (muscular weakness) occurred more frequently in patients assigned to only the statin drug compared to patients taking both the statin and niacin.
Even though significant between-group differences occurred in the numbers of serious adverse events, overall there wasn't much of a difference between the groups. According to the study, "Overall, 34.2 percent of patients who received extended-release niacin and 32.5 percent of patients who received placebo had serious adverse events during follow-up."
The study authors, in a supplement appendix to their letter to the editor of The New England Journal of Medicine, urged that the adverse event data be interpreted with caution. They pointed out that the formulation of niacin used in the AIM-HIGH trial has been used clinically since 1997. Niacin also has been used to treat imbalanced lipid levels since the 1950s. The study authors wrote, "Thus, information concerning the utility and safety profile of this agent has been acquired during almost 60 years of clinical use. Numerous trials and review articles have reported on the adverse effects related to niacin in both the immediate-release and extended-release formulations, including that used in AIM-HIGH. Thus, adverse events, as determined by this trial of 3,414 subjects, but not previously described with niacin, should be considered in the context of the long prior clinical and trial experience."
The second study mentioned in The New York Times blog is the Heart Protection Study 2: Treatment of HDL to Reduce the Incidence of Vascular Events (HPS2-THRIVE). In this study, researchers randomly divided 25,673 subjects with vascular disease or diabetes into two groups. One group received 2 grams daily of extended-release niacin and 40 mg of laropiprant, a drug that antagonizes the prostaglandin D2 receptor DP1 and which has been shown to reduce niacin flushing. The other group received a placebo. The patients also were given the statin drug simvastatin and, when needed, the drug ezetimibe.3
The results of this study were similar to the AIM-HIGH trial. Niacin did not reduce the risk of major vascular events (coronary events or stroke). Niacin did produce a "nominally significant 10 percent proportional reduction in arterial revascularization procedures."
Compared to subjects in the placebo group, the subjects who took the niacin–laropiprant combination had significantly more fatal or nonfatal serious adverse events (7,137 vs. 6,762). The most common adverse effect was disturbances in glucose metabolism. When the researchers analyzed the 8,299 diabetic subjects, compared to placebo, niacin–laropiprant was linked to "a 55 percent proportional increase in disturbances in diabetes control that were considered to be serious, most of which led to hospitalization."3
In the 17,374 non-diabetic participants, niacin–laropiprant resulted in a "32 percent proportional increase in diagnosis of diabetes" compared with placebo.
Similar to the AIM-HIGH trial, there were increased infections in subjects given extended-release niacin and laropiprant. There was also an increase in bleeding in many of the subjects taking niacin-laropiprant.
The primary thing to keep in mind about the HPS2-THRIVE study is that it really isn't about niacin. The study is about a niacin-laropiprant combination. The investigators did not compare the effects of niacin-laropiprant to the effects of niacin alone. Laropiprant is a selective prostaglandin-2 receptor inhibitor that was rejected in 2008 by the FDA for marketing because of its side effect profile. Laropiprant is metabolized in the liver, which begs the question of how much of an overall physiological burden it is placing on the body.
This trial also studied more than 25,000 participants. In a study with 25,000-plus people, it is possible to find statistically significant differences that are practically meaningless.
Niacin's effects on blood glucose are nothing new. It is widely known that it can have this effect. However, other researchers have noted that niacin's beneficial effects on lipids may outweigh this side effect.
Writing in the Mayo Clinic Proceedings, one group of reviewers noted, "On the basis of our analysis, the effects of niacin (< or =2.5 grams/day), alone or in combination with statins, on fasting glucose (an increase of 4 percent-5 percent) and hemoglobin A1c levels (an increase of < or =0.3 percent) are modest, transient, or reversible, and typically amenable to adjustments in oral hypoglycemic regimens without discontinuing niacin. Niacin therapy was infrequently associated with incident diabetes or the need for new insulin prescriptions. Studies showed important clinical benefits of niacin or niacin-statin regimens despite modest effects on glucose control. On a population basis, significant reductions in incidences of cardiovascular events and the degree of atherosclerotic progression associated with long-term niacin (or niacin-statin) therapy in patients with diabetic dyslipidemia outweigh the typically mild effects of this therapy on glycemic regulation. Consensus guidelines recommend monitoring glycemic control after initiating niacin treatment or increasing its dosage."4
Additional Considerations to Keep in Mind
The AIM-HIGH and HPS2-THRIVE studies failed to examine other medications the subjects were taking, dietary and drug-drug interactions, the effects of hormone therapy, and other considerations that should be factored into such studies in order to obtain the most accurate results. The studies also did not factor in the particle size of HDL or LDL.
People who have small particle size LDL are at a substantially higher risk than an individual with the same quantity of LDL of large particle size.
Additionally, the amount of niacin used in these two studies was very high. With the exception of people who have genetic abnormalities in the way they process lipids, most patients don't need to use such high levels of niacin when they are also incorporating and adopting a full wellness program
The bottom line: In the hands of a skilled functional medicine provider, niacin is a solid tool, when combined with good nutrition, exercise, and other dietary and supplemental considerations.
- Boden WE, et al. N Engl J Med. 2011 Dec 15;365(24):2255-67.
- Anderson T, et al. N Engl J Med. July 2014; 371:288-90.
- Landray MJ, et al. N Engl J Med. July 17, 2014;371(3):203-12.
- Goldberg RB and Jacobson TA. Mayo Clin Proc. 2008 Apr;83(4):470-8.