Niacin, Cardiovascular disease, Cochrane analysis

Statistical Significance Does Not Always Mean Clinical Relevance

By George Milowe, MD

Well the number crunchers and statisticians are at it again. While I sadly read of the demise of niacin for the prevention of cardiovascular events in the latest Cochrane Library analysis, I was not swayed by their data “pooling” or meta-analysis. What Cochrane did was to pool large studies over the years in order to determine if niacin by itself was efficacious or if, when added to other drugs such as statins, it increased their efficacy in the prevention of cardiovascular disease. At the end of all the number crunching their conclusion was “moderate to high quality evidence suggests niacin does not reduce mortality, cardiovascular mortality, non-cardiovascular mortality, the number of fatal or non-fight fatal myocardial infarctions, or the number of fatal or non-fatal strokes but is associated with side effects. Benefits from niacin therapy in the prevention of cardiovascular disease events are unlikely.”¹

The question that I have is did they properly evaluate the clinical information in many of these studies? For example, the median duration of their 23 randomized controlled studies with a total of 39,125 participants, was 11.5 months and the median dose of niacin was only 2 grams a day. The two largest studies used by Cochrane that skewed the results were the Atherothrombosis Intervention in Metabolic Syndrome With Low HDL/High Triglycerides and Impact on Global Health Outcomes (AIM-HIGH),² 3400 participants, and Heart Protection Study 2–Treatment of HDL to Reduce the Incidence of Vascular Events (HPS2-THRIVE),³ 25,000 participants. These two large studies used 1.9 grams on average daily of the niacin ER (extended release) form. First, the maximum daily recommended dose of the ER form is only 2 g while it is known that immediate release (IR) niacin is more effective and can be given in higher doses. Better results are achieved with 3 to 4 g or more (up to 6 g) of IR daily. Second, the selection of patients is crucial.

The patients that do best with niacin are those that have metabolic syndrome (MS). And yes, that would be patients who have high to very high triglyceride levels and very low HDL. In the HPS2-THRIVE and AIM-HIGH studies, baseline triglyceride (TG) levels were relatively modest or mildly elevated: 127mg/dl (HPS2-THRIVE) and 161mg/dl (AIM HIGH). The HPS2-THRIVE study alone comprised about 73% of Cochrane’s entire study population. The current TG criterion for MS is 150 or higher. The HPS2-THRIVE group at study entry treated with a statin had low LDL (63) and relatively high HDL (49.6) This is a group unlikely to be helped further by niacin. The AIM HIGH at entry had LDL 71, HDL 34.9. so possibly a group to be further helped with niacin. In fact, a sub-analysis in this group of those patients with high TG (198 mg/dl or greater) and lower LDL at entry (less than 33 mg/dl) showed a potential benefit of combined statin and niacin therapy with a hazard ratio of 0.74 (p=.07).⁴

It is a well-known fact that higher triglyceride levels cause higher LDL particle counts with higher concentrations of small dense LDL, which are more atherogenic then the larger less dense LDL particles.⁵ The Cochrane analysis study was swayed mostly by these two very large studies, which did not have optimal patient selection for the addition of niacin to a statin. Also, optimal doses of niacin were not used. A good paper which brings these points to bear was published just a few months after the Cochrane analysis and should be read by every practitioner.⁴ These authors also point out that the IR niacin should not be given at bedtime but rather at mealtime. At bedtime when a patient is fasting or relatively fasting, it induces a sympathetic response and that also can provoke cardiovascular events. Finally, the authors also point out that the best absolute mortality benefit of any statin trial was in the Scandinavian Simvastatin Survival Study (SSSS),⁵ producing an absolute mortality benefit (AMB) of 3.5% in established coronary artery disease over 5-6 years (P=.0003). For niacin studies, the 15-year follow-up in the Coronary Drug Project⁶ yielded a higher AMB of 6.2% (P=.0004) while the Stockholm Ischemic Heart Disease⁷ showed a 7.8% (P=.035) AMB using niacin plus clofibrate over five years vs. no lipid medications in the placebo group. These latter niacin trials were from 1986 and 1988 and were not included in the Cochrane analysis!

It is true that some people have difficulty tolerating niacin, but many people do get used to the flushing and it can wane with time. There is a tendency of an increased risk for diabetes with niacin as with statins. But overall, the side effect profile is much better than with statins.⁸ There are no muscle aches and pains, myopathies, neuropathies, memory problems, or lower steroid hormone levels that you’ll see with the statins. Long term statin use can cause impairment of cardiac diastolic function^9,10 (likely due to reduction of synthesis of ubiquinone and Heme-A) and increased risk of calcification of arterial walls and connective tissue due to blockage of vitamin K2 synthesis.⁹ Niacin also has anti-inflammatory properties and is also being studied in other medical conditions such as inflammatory bowel disease.¹¹ So niacin for cardiovascular patients is not dead, and there are many practitioners out there who are smart enough to still be using it with properly selected patients.

References

1. Cochrane Database Syst Rev. 2017 Jun; 2017(6): CD009744

2. N Engl J Med. 2011; 365; 2255-226.

3. N Engl J Med. 2014;371:203-212.

4. J Clin Lipidol. Nov-Dec 2017;11(6):1309-1317.

5. Lancet. 1994 Nov 19;344(8934):1383-9.

6. J Am Coll Cardiol. 986;8:1245-1255.

7. Acta Med Scand. 1988;223:405-418.

8. Am J Cardiol. 2007 Mar 19;99(6A):22C-31C.

9. Expert Rev Clin Pharmacol. 2015 Mar;8(2):189-99.

10. Perm J. 2019;23:18.257.

11. EMBO Mol Med. 2017 May; 9(5): 571–588.