Zissimopoulos JM, Barthold D, Brinton RD, Joyce G.
Sex and Race Differences in the Association Between Statin Use and the Incidence of Alzheimer Disease.
JAMA Neurol. 2017 Feb 1;74(2):225-232.
PubMed.
This appears to be a carefully conceived and conducted prospective observational study of the relation between low- vs. higher-usage of the common statin drugs and incident Alzheimer’s dementia (AD). The study incorporates some clever design features, e.g., establishing exposure classification before the risk period for the outcome of AD diagnosis.
It exemplifies the usual strengths and limitations of large-scale, database-driven research of this kind. Numbers are large, inviting a comparison of the “effect” of high vs. low statin exposure for all the commonly used statin medicines, and also permitting sub-analyses of “effect” segregated by sex and a number of racial and ethnic categories.
But we must bear in mind that a larger observational study is no less vulnerable to systematic bias or confounding than a smaller one. Here, the apparent “effect” of statin exposure is relatively small, about 10 percent lower incidence of AD among the higher statin use groups. While still important from a public health perspective, this sort of risk modification can be particularly vulnerable in observational studies to subtle residual confounding from many sources, and it should therefore be interpreted with circumspection.
The differences in AD incidence by ethnic groups recapitulate earlier findings, but again, because the differences in rates are quite small, it is difficult to interpret these differences either by total or by individual agent statin usage.
The data do show that AD rates are higher in nonusers of statins than in either the low-use or high-use groups. This tends to support the idea that statin use is protective against AD, but the authors are careful to note the many possible sources of confounding (especially confounding by indication) in this last comparison.
One puzzling finding is the lack of much if any difference in statin “effects” between lipophilic vs. hydrophilic agents; one might have expected lipophilic agents to be more potent as AD preventives because of their likely blood-brain barrier penetration. This point is obscured somewhat, however, by the fact that hydrophilic agents were much less commonly used in the exposure interval (2006-08) than they are today, resulting in wider confidence intervals in their analysis.
The relatively small risk reduction reported here in the high- vs. low-use groups may explain some of the difficulties encountered by the randomized trials that have been carried out to test the statin prevention hypothesis. Vast numbers of individuals with lengthy follow-up would be needed to accumulate sufficient statistical power to demonstrate an effect of this size in a condition such as AD with an incidence rate near 1 or 2 percent. This paper may therefore help clarify some of the conflicting literature about the statin—AD relationship and, in particular, the failure of several trials to show a significant benefit of these drugs.
I am afraid we won’t be able to establish more definitively the effect of statins using trials. With an effect size so small, it would require huge numbers of enrollees and cost tens of millions of dollars. I doubt anyone will sponsor that, given the vulnerabilities of the methodology to date.
This is an interesting article. We have demonstrated that peroxisome proliferator-activated receptor alpha (PPARα) is constitutively present in the hippocampus (Roy et al., 2013; Corbett et al., 2015), and that PPARα serves as a receptor for statins for its non-canonical functions, such as neurotrophin synthesis (Roy et al., 2015).
Since neurotrophins play an important role in the Alzheimer’s disease process, and since we have also shown that simvastatin is unable to protect memory and learning in an AD animal models lacking PPARα (Roy et al., 2015), it is therefore possible that patients who see no benefit from statins may have less PPARα in their hippocampi.
References:
Roy A, Jana M, Corbett GT, Ramaswamy S, Kordower JH, Gonzalez FJ, Pahan K.
Regulation of cyclic AMP response element binding and hippocampal plasticity-related genes by peroxisome proliferator-activated receptor α.
Cell Rep. 2013 Aug 29;4(4):724-37. Epub 2013 Aug 22
PubMed.
Corbett GT, Gonzalez FJ, Pahan K.
Activation of peroxisome proliferator-activated receptor α stimulates ADAM10-mediated proteolysis of APP.
Proc Natl Acad Sci U S A. 2015 Jul 7;112(27):8445-50. Epub 2015 Jun 15
PubMed.
Roy A, Jana M, Kundu M, Corbett GT, Rangaswamy SB, Mishra RK, Luan CH, Gonzalez FJ, Pahan K.
HMG-CoA Reductase Inhibitors Bind to PPARα to Upregulate Neurotrophin Expression in the Brain and Improve Memory in Mice.
Cell Metab. 2015 Aug 4;22(2):253-65. Epub 2015 Jun 25
PubMed.
This somewhat disappointing report does not adequately take into account decades of research. The clinical trials recommended by the authors have already been conducted, and they have quite thoroughly answered the question of whether statins reduce dementia and cognitive loss. They do not. Further, the paper highlights the continuing lack of understanding of the difference between association and causation.
The clinical trials needed to address these issues have been done and published. Specifically, the administrative data set published here reflects the use of statins most likely for the treatment of hypercholesterolemia and risk of cardiovascular disease. This data set does not enable a determination of whether the untreated actually should have been treated because they had risk. This is in part because information such as cholesterol levels is unavailable.
The study did find a reduced risk of AD with high use of simvastatin /therapeutics/simvastatin (as well as the other statins), and the authors suggested that a similar reduction may not have been seen in previous studies because people with dyslipidemia, who may have most benefited from statins, were excluded. However, there is an ethical dilemma in conducting placebo-controlled trials that require withholding an approved treatment from people who should receive the medication for a different indication.
Others have conducted studies of patients with dyslipidemia that measured cognition, and they found great cardiovascular protection without an effect on cognition. These studies included thousands of people and were highly published years ago (see Heart Protection Study Collaborative Group, 2002; Shepherd et al., 2002).
The present study reinforces the general medical advice of treating a person’s treatable conditions. In this case, if you have dyslipidemia or other risk factors for poor cardiovascular outcomes, treat those conditions. I suspect that if you are not treating those conditions, you are medically underserved, and being medically underserved is a source of bad medical outcomes across the board, dementia among them. In essence, we are not talking about a drug effect of statins on dementia as much as about an effect of inadequate care on multiple bad outcomes.
References:
MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial.
Lancet. 2002 Jul 6;360(9326):7-22.
PubMed.
Shepherd J, Blauw GJ, Murphy MB, Bollen EL, Buckley BM, Cobbe SM, Ford I, Gaw A, Hyland M, Jukema JW, Kamper AM, Macfarlane PW, Meinders AE, Norrie J, Packard CJ, Perry IJ, Stott DJ, Sweeney BJ, Twomey C, Westendorp RG, PROSPER study group. PROspective Study of Pravastatin in the Elderly at Risk.
Pravastatin in elderly individuals at risk of vascular disease (PROSPER): a randomised controlled trial.
Lancet. 2002 Nov 23;360(9346):1623-30.
PubMed.
This interesting article provides additional hope that the cholesterol-lowering statins may help reduce the risk of AD. Unfortunately, as others pointed out, results from clinical trials with statins for AD have been disappointing.
However, it is worth noting that as potent inhibitors of HMG-CoA reductase in the mevalonate pathway, statins also inhibit the production of isoprenoid intermediates in addition to the final product cholesterol. The short-chain lipid isoprenoids, mainly farnesyl pyrophosphate (FPP) and geranylgeranyl PP (GGPP), serve as lipid donors for a post-translational process called protein prenylation, catalyzed by farnesyltransferase and geranylgeranyltransferases, respectively. The isoprenyl groups facilitate anchoring of proteins in cell membranes and mediate protein-protein interactions. Prenylated proteins, including small GTPases comprising the Ras superfamily along with heterotrimeric G-proteins, are involved in regulating diverse cellular processes, including synaptic and cognitive function (Hottman and Li, 2014).
Many of the pleiotropic effects of statins have been shown to be mediated through the reduction of isoprenoids and inhibition of protein prenylation. Importantly, we have demonstrated that the two protein prenylation pathways play distinct roles in AD (Cheng et al., 2013). Reducing protein farnesylation rescued cognitive function as well as attenuated amyloid pathology and neuroinflammation whereas reducing protein geranylgeranylation failed to produce cognitive benefits in a mouse model of AD (Cheng et al., 2013). Depending on the type, dose, and treatment duration of statins, the two prenylation pathways can be affected to different degrees (Winter-Vann and Casey, 2005), which may lead to inconsistent outcomes.
The role of isoprenoids, protein prenylation, and related downstream signaling pathways in AD is markedly underexplored but could be extremely important mechanistically and therapeutically. The levels of FPP and GGPP are increased in brains of patients with AD. This could potentially disrupt the give-and-take balance between prenylated Ras, RhoA and Rac1, impacting synaptic plasticity (Eckert et al., 2009).
Our recent studies with human brain tissue show that dysregulation of protein prenylation occurs in early stages of AD and is associated with cognitive impairment and neuropathology (unpublished). A better understanding of the distinct roles of the two protein prenylation pathways in AD is necessary in order to develop more specific and effective therapeutics for AD beyond statins.
References:
Cheng S, Cao D, Hottman DA, Yuan L, Bergo MO, Li L.
Farnesyl Transferase Haplodeficiency Reduces Neuropathology and Rescues Cognitive Function in a Mouse Model of Alzheimer's Disease.
J Biol Chem. 2013 Oct 17;
PubMed.
Eckert GP, Hooff GP, Strandjord DM, Igbavboa U, Volmer DA, Müller WE, Wood WG.
Regulation of the brain isoprenoids farnesyl- and geranylgeranylpyrophosphate is altered in male Alzheimer patients.
Neurobiol Dis. 2009 Aug;35(2):251-7.
PubMed.
Hottman DA, Li L.
Protein Prenylation and Synaptic Plasticity: Implications for Alzheimer's Disease.
Mol Neurobiol. 2014 Jan 5;
PubMed.
Winter-Vann AM, Casey PJ.
Post-prenylation-processing enzymes as new targets in oncogenesis.
Nat Rev Cancer. 2005 May;5(5):405-12.
PubMed.
Comments
McGill University Faculty of Medicine
This appears to be a carefully conceived and conducted prospective observational study of the relation between low- vs. higher-usage of the common statin drugs and incident Alzheimer’s dementia (AD). The study incorporates some clever design features, e.g., establishing exposure classification before the risk period for the outcome of AD diagnosis.
It exemplifies the usual strengths and limitations of large-scale, database-driven research of this kind. Numbers are large, inviting a comparison of the “effect” of high vs. low statin exposure for all the commonly used statin medicines, and also permitting sub-analyses of “effect” segregated by sex and a number of racial and ethnic categories.
But we must bear in mind that a larger observational study is no less vulnerable to systematic bias or confounding than a smaller one. Here, the apparent “effect” of statin exposure is relatively small, about 10 percent lower incidence of AD among the higher statin use groups. While still important from a public health perspective, this sort of risk modification can be particularly vulnerable in observational studies to subtle residual confounding from many sources, and it should therefore be interpreted with circumspection.
The differences in AD incidence by ethnic groups recapitulate earlier findings, but again, because the differences in rates are quite small, it is difficult to interpret these differences either by total or by individual agent statin usage.
The data do show that AD rates are higher in nonusers of statins than in either the low-use or high-use groups. This tends to support the idea that statin use is protective against AD, but the authors are careful to note the many possible sources of confounding (especially confounding by indication) in this last comparison.
One puzzling finding is the lack of much if any difference in statin “effects” between lipophilic vs. hydrophilic agents; one might have expected lipophilic agents to be more potent as AD preventives because of their likely blood-brain barrier penetration. This point is obscured somewhat, however, by the fact that hydrophilic agents were much less commonly used in the exposure interval (2006-08) than they are today, resulting in wider confidence intervals in their analysis.
The relatively small risk reduction reported here in the high- vs. low-use groups may explain some of the difficulties encountered by the randomized trials that have been carried out to test the statin prevention hypothesis. Vast numbers of individuals with lengthy follow-up would be needed to accumulate sufficient statistical power to demonstrate an effect of this size in a condition such as AD with an incidence rate near 1 or 2 percent. This paper may therefore help clarify some of the conflicting literature about the statin—AD relationship and, in particular, the failure of several trials to show a significant benefit of these drugs.
I am afraid we won’t be able to establish more definitively the effect of statins using trials. With an effect size so small, it would require huge numbers of enrollees and cost tens of millions of dollars. I doubt anyone will sponsor that, given the vulnerabilities of the methodology to date.
View all comments by John BreitnerFloyd A. Davis, M.D., Endowed Chair in Neurology Rush University Medical Center
This is an interesting article. We have demonstrated that peroxisome proliferator-activated receptor alpha (PPARα) is constitutively present in the hippocampus (Roy et al., 2013; Corbett et al., 2015), and that PPARα serves as a receptor for statins for its non-canonical functions, such as neurotrophin synthesis (Roy et al., 2015).
Since neurotrophins play an important role in the Alzheimer’s disease process, and since we have also shown that simvastatin is unable to protect memory and learning in an AD animal models lacking PPARα (Roy et al., 2015), it is therefore possible that patients who see no benefit from statins may have less PPARα in their hippocampi.
References:
Roy A, Jana M, Corbett GT, Ramaswamy S, Kordower JH, Gonzalez FJ, Pahan K. Regulation of cyclic AMP response element binding and hippocampal plasticity-related genes by peroxisome proliferator-activated receptor α. Cell Rep. 2013 Aug 29;4(4):724-37. Epub 2013 Aug 22 PubMed.
Corbett GT, Gonzalez FJ, Pahan K. Activation of peroxisome proliferator-activated receptor α stimulates ADAM10-mediated proteolysis of APP. Proc Natl Acad Sci U S A. 2015 Jul 7;112(27):8445-50. Epub 2015 Jun 15 PubMed.
Roy A, Jana M, Kundu M, Corbett GT, Rangaswamy SB, Mishra RK, Luan CH, Gonzalez FJ, Pahan K. HMG-CoA Reductase Inhibitors Bind to PPARα to Upregulate Neurotrophin Expression in the Brain and Improve Memory in Mice. Cell Metab. 2015 Aug 4;22(2):253-65. Epub 2015 Jun 25 PubMed.
View all comments by Kalipada PahanMount Sinai School of Medicine
This somewhat disappointing report does not adequately take into account decades of research. The clinical trials recommended by the authors have already been conducted, and they have quite thoroughly answered the question of whether statins reduce dementia and cognitive loss. They do not. Further, the paper highlights the continuing lack of understanding of the difference between association and causation.
The clinical trials needed to address these issues have been done and published. Specifically, the administrative data set published here reflects the use of statins most likely for the treatment of hypercholesterolemia and risk of cardiovascular disease. This data set does not enable a determination of whether the untreated actually should have been treated because they had risk. This is in part because information such as cholesterol levels is unavailable.
The study did find a reduced risk of AD with high use of simvastatin /therapeutics/simvastatin (as well as the other statins), and the authors suggested that a similar reduction may not have been seen in previous studies because people with dyslipidemia, who may have most benefited from statins, were excluded. However, there is an ethical dilemma in conducting placebo-controlled trials that require withholding an approved treatment from people who should receive the medication for a different indication.
Others have conducted studies of patients with dyslipidemia that measured cognition, and they found great cardiovascular protection without an effect on cognition. These studies included thousands of people and were highly published years ago (see Heart Protection Study Collaborative Group, 2002; Shepherd et al., 2002).
The present study reinforces the general medical advice of treating a person’s treatable conditions. In this case, if you have dyslipidemia or other risk factors for poor cardiovascular outcomes, treat those conditions. I suspect that if you are not treating those conditions, you are medically underserved, and being medically underserved is a source of bad medical outcomes across the board, dementia among them. In essence, we are not talking about a drug effect of statins on dementia as much as about an effect of inadequate care on multiple bad outcomes.
References:
MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet. 2002 Jul 6;360(9326):7-22. PubMed.
Shepherd J, Blauw GJ, Murphy MB, Bollen EL, Buckley BM, Cobbe SM, Ford I, Gaw A, Hyland M, Jukema JW, Kamper AM, Macfarlane PW, Meinders AE, Norrie J, Packard CJ, Perry IJ, Stott DJ, Sweeney BJ, Twomey C, Westendorp RG, PROSPER study group. PROspective Study of Pravastatin in the Elderly at Risk. Pravastatin in elderly individuals at risk of vascular disease (PROSPER): a randomised controlled trial. Lancet. 2002 Nov 23;360(9346):1623-30. PubMed.
View all comments by Mary SanoUniversity of Minnesota, Twin Cities
This interesting article provides additional hope that the cholesterol-lowering statins may help reduce the risk of AD. Unfortunately, as others pointed out, results from clinical trials with statins for AD have been disappointing.
However, it is worth noting that as potent inhibitors of HMG-CoA reductase in the mevalonate pathway, statins also inhibit the production of isoprenoid intermediates in addition to the final product cholesterol. The short-chain lipid isoprenoids, mainly farnesyl pyrophosphate (FPP) and geranylgeranyl PP (GGPP), serve as lipid donors for a post-translational process called protein prenylation, catalyzed by farnesyltransferase and geranylgeranyltransferases, respectively. The isoprenyl groups facilitate anchoring of proteins in cell membranes and mediate protein-protein interactions. Prenylated proteins, including small GTPases comprising the Ras superfamily along with heterotrimeric G-proteins, are involved in regulating diverse cellular processes, including synaptic and cognitive function (Hottman and Li, 2014).
Many of the pleiotropic effects of statins have been shown to be mediated through the reduction of isoprenoids and inhibition of protein prenylation. Importantly, we have demonstrated that the two protein prenylation pathways play distinct roles in AD (Cheng et al., 2013). Reducing protein farnesylation rescued cognitive function as well as attenuated amyloid pathology and neuroinflammation whereas reducing protein geranylgeranylation failed to produce cognitive benefits in a mouse model of AD (Cheng et al., 2013). Depending on the type, dose, and treatment duration of statins, the two prenylation pathways can be affected to different degrees (Winter-Vann and Casey, 2005), which may lead to inconsistent outcomes.
The role of isoprenoids, protein prenylation, and related downstream signaling pathways in AD is markedly underexplored but could be extremely important mechanistically and therapeutically. The levels of FPP and GGPP are increased in brains of patients with AD. This could potentially disrupt the give-and-take balance between prenylated Ras, RhoA and Rac1, impacting synaptic plasticity (Eckert et al., 2009).
Our recent studies with human brain tissue show that dysregulation of protein prenylation occurs in early stages of AD and is associated with cognitive impairment and neuropathology (unpublished). A better understanding of the distinct roles of the two protein prenylation pathways in AD is necessary in order to develop more specific and effective therapeutics for AD beyond statins.
References:
Cheng S, Cao D, Hottman DA, Yuan L, Bergo MO, Li L. Farnesyl Transferase Haplodeficiency Reduces Neuropathology and Rescues Cognitive Function in a Mouse Model of Alzheimer's Disease. J Biol Chem. 2013 Oct 17; PubMed.
Eckert GP, Hooff GP, Strandjord DM, Igbavboa U, Volmer DA, Müller WE, Wood WG. Regulation of the brain isoprenoids farnesyl- and geranylgeranylpyrophosphate is altered in male Alzheimer patients. Neurobiol Dis. 2009 Aug;35(2):251-7. PubMed.
Hottman DA, Li L. Protein Prenylation and Synaptic Plasticity: Implications for Alzheimer's Disease. Mol Neurobiol. 2014 Jan 5; PubMed.
Winter-Vann AM, Casey PJ. Post-prenylation-processing enzymes as new targets in oncogenesis. Nat Rev Cancer. 2005 May;5(5):405-12. PubMed.
View all comments by Ling LiMake a Comment
To make a comment you must login or register.