Aspirin in elderly increases anemia

 A secondary analysis of the ASPREE study found an increase in anemia in the elderly taking low-dose aspirin (see aspirin inc anemia AnnIntMed2023 in dropbox or  doi:10.7326/M23-0675). 

 

Details:

-- 18,153 persons were randomly assigned to 100 mg of aspirin daily or placebo in the ASPirin in Reducing Events in the Elderly (ASPREE) trial of community-dwelling persons at least 70 years old, or Black or Hispanic persons at least 65 years old, in Australia and the US, with enrollment from 2010 to 2015

-- mean age 74, 56% female, 87% were white from Australia, 6% were white from the US, 3.9% Black, 2.5% Hispanic

-- all were community-based: 32% lived at home alone, 68% with others

-- smoking status was never in 55%/former in 41%/current in 4%; cancer history 19%; BMI 24% normal/29% obese/45% overweight; waist circumference high in 45%

-- mean baseline hemoglobin 14.2 g/dL

-- mean blood pressure 139/77, eGFR 74, CKD 29%, hypertension 74%, use of PPI 24%

 

-- hemoglobin levels were measured annually

 

-- primary outcome: incident anemia in the elderly as defined by the WHO: hemoglobin <12 g/dL for men and <11 g/dL for women

-- prespecified secondary outcome: major hemorrhage, defined as a composite of hemorrhagic stroke, symptomatic intracranial bleeding, or clinically significant extracranial bleeding (bleeding that led to transfusion, hospitalization, prolongation of hospitalization, surgery, or death)

-- other secondary analyses included ferritin concentrations and hemoglobin concentrations

 

-- median duration follow-up: 4.7 years (3.6-5.7 years); median number of hemoglobin measurements = 5

 

Results:

-- anemia incidence:

    -- aspirin group: 51.2 events per 1000 person-years

    -- placebo: 42.9 events per 1000 person-years

        -- 20% increase in those on aspirin, HR 1.20 (1.12-1.29)

    -- estimated probability of experiencing anemia within five years:

        -- aspirin group: 23.5% (22.4%-24.6%)

        -- placebo group: 20.3% (19.3%-21.4%)

    -- these results are consistent across subgroups by age, sex, chronic kidney disease, diabetes, smoking status, alcohol use, prior aspirin use, and use of an NSAID or proton pump inhibitors

 

-- Hemoglobin concentrations:

    -- aspirin group: steeper decline than the placebo group by 0.6 g/L at five years (0.3-1.0 g/L), where the placebo group decreased 3.6 g/L at five years

 

-- ferritin concentrations <45 mcg/L (measured in 7139 participants at baseline and at year three):

    -- aspirin group: 465 participants (13%)

    -- placebo: 350 participants (9.8%)

    -- aspirin group versus placebo: overall decline in ferritin by 11.5% (9.3%-13.7%)

-- ferritin concentrations <100 mcg/L (measured in 7139 participants at baseline and at year three):

    -- aspirin group: 1395 participants (39%)

    -- placebo: 1116 participants (31%)

    -- aspirin group versus placebo: overall decline in ferritin by 11.5% (9.3%-13.7%)

 

-- although all of the above numbers reflected small absolute changes in these parameters, all of the graphs showed increasing differences over time, in a pretty linear manner

 

-- sensitivity analyses:

    -- not including major bleeding: similar results

    -- not including cancer incidence or other characteristics associated with increased risk of anemia: similar results

 

-- Clinically significant bleeding:

    -- aspirin group: 273 participants (3.0%)

    -- placebo: 192 (2.1%)

 

Commentary:

-- world-wide prevalence of anemia in people >75yo is about 30%; in community-based studies it is approximately 12%

-- in the elderly, anemia is most commonly associated with iron deficiency (15%-20%), medical comorbidities (including kidney disease), and/or inflammation (especially anemia of chronic disease)

-- anemia is associated with functional deficits, morbidity, and mortality in the elderly

-- approximately 50% of persons over 60 years old in the United States reported taking aspirin preventatively from 2011 to 2018

-- several studies have found that low-dose aspirin taken daily increases major bleeding complications (as also found in the above study).

 

-- This was the first large study to assess incident anemia in those without major bleeds, documenting:

    -- the rate of incident anemia was about 20% higher in those on low-dose aspirin, along with a decrease in mean hemoglobin and ferritin concentrations

    -- there was a progressive decrease in hemoglobin levels accompanying this increased incidence of anemia vs placebo:  graphs for both hemoglobin levels and anemia had  linear decreases that progressively widened for those on aspirin over the five year period

        -- since aspirin is an appropriate very long-term intervention for many individuals as they get older and have more serious cardiovascular events, the likelihood of having significant occult and severe bleeding over potentially decades of aspirin would likely increase significantly

 

-- One concerning issue is the use of aspirin for primary prevention of cardiovascular disease:

    -- In April 2022, the US Preventive Services Task Force downgraded its recommendation for the use of aspirin in primary cardiovascular prevention to “grade C” for those 40 to 59 years old with at least a 10% cardiovascular risk, and “grade D” for those over age 60, where grade C means no recommendation and grade D means don’t do it

    -- in 2011-2019, 45.6% of those >70yo without cardiovasc disease took aspirin (https://pubmed.ncbi.nlm.nih.gov/34632437/ )

        -- this did not include those people buying aspirin on their own over-the-counter

        -- and, many elderly may be continuing with the OTC aspirin, even if their clinicians did not suggest it, suggested stopping it, or didn’t read the USPSTF recommendation changes….

    -- for a critique of using the ASCVD risk assessment tools especially in those 40-59yo, as per USPSTF recommendations, see https://gmodestmedblogs.blogspot.com/2023/10/update-ascvd-risk-factor-critique.html

-- Mechanistically, aspirin is associated with inhibition of platelet aggregation, predisposing individuals to an increased likelihood of having a bleeding event, as well as inhibition of the COX-1 enzyme, which adds to the issue by reducing gastric prostaglandin production, thereby leading to a reduction of the protective barrier for the gut wall

 

-- studies have found that enteric-coated aspirin might be less effective in platelet inhibition that non-coated aspirin, though one recent large study (secondary analysis of the ADAPTABLE RCT) found that in 15,000 patients with atherosclerotic disease did not find a significant difference in effectiveness on major clinical cardiovascular outcomes if taking the enteric-coated aspirin: cad aspirin regular vs enteric-coated JAMAcardiol2023 in drobbox, or https://jamanetwork.com/journals/jamacardiology/fullarticle/2809795 . However, there was no difference in bleeding between those on enteric-coated versus regular aspirin. This lack of GI protection by taking enteric-coated aspirin has been found in several other studies (eg, as in a 2021 systematic review: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8403009/).

 

-- there is some controversy as to what the upper limit of ferritin is in diagnosing iron deficiency with or without anemia. The frequently held-belief is that ferritin levels between 40 and 200 are normal. However, some studies have found completely depleted bone marrow iron stores with ferritin as high as 100, and this may be especially true in those with chronic diseases such as heart failure, CKD, inflammatory bowel disease: https://www.hindawi.com/journals/ijcd/2018/9394060/ . so, perhaps the numbers above using the ferritin <100 mcg/L is more appropriate

-- One tangential issue (which I allude to periodically since it is so important for us in dosing medications, etc) is that we usually assess creatinine concentration to derive the calculated eGFR, but creatinine seem to  have a poor correlation with actual measured GFR (https://gmodestmedblogs.blogspot.com/2022/07/egfr-not-such-great-estimate-of-renal.html ), and this is especially true in the elderly who are more likely to have decreased muscle mass (eg normal creatinine, but severe CKD by cystatin c, as found in several of my patients), though the accuracy of cystatin c is also suspect

 

Limitations:

-- this was a post-hoc analysis of a large trial, limiting its statistical robustness

-- hemoglobins were measured annually, and there was the potential issue that in the course of the year, some patients had received transfusion or therapy for iron deficiency

-- there is not much ethnic diversity in the study (93% were white) and it involved predominantly people from Australia (88%), both of which could limit generalizability of their results

    -- especially since taking vitamins with iron, or eating food with higher iron content, may vary widely in different areas and different cultures

-- there was no robust analysis of the cause of anemia in this study, but the strong likelihood was it was from bleeding/iron deficiency  given the attendant decrease in ferritin levels.

    -- Though, it should be appreciated that ferritin is not a specific indicator of iron stores, since it can be elevated in inflammatory conditions, though low ferritin is pretty specific for iron deficiency

-- it is difficult to know the real numbers of people taking aspirin, given its widespread availability over-the-counter

-- 24% of people were on PPIs, which might limit the development of anemia and understate the risk of aspirin-induced anemia, though PPIs also decrease iron absorption in the stomach

-- though they included many baseline characteristics, many involved binary cutpoints (eg waist circumference, smoking history, hypertension, chronic kidney disease).  This undercuts the more nuanced (and helpful) understanding of the individual patients we see, since CKD can be a creatinine of 1.1 or 13.1; etc

    -- and many of their outcomes are binary, and negate the continuum of risk: is there really a difference when a male, for instance, has a hemoblobin of 11.9 g/dL or 12.0 g/dL??

-- this was a five-year study, and many individuals are appropriately on aspirin for many more years. The cumulative incidence of anemia in the study increased linearly from the first to the fifth year and likely would continue increasing in patients who will be on the aspirin for a longer period of time, though this study could not prove that

-- the identification of serious bleeding events is limited to those treated at a hospital, and did not include patients who had severe epistaxis, for example, which may have been treated in an outpatient setting

 

So, this study does bring up a few issues:

-- aspirin has clearly been associated with major GI bleeding in many studies, and it certainly makes sense that it would be associated with occult bleeding as found here

-- it is clear that people with significant cardiovascular disease should be on aspirin, though many not in this secondary prevention category are likely taking aspirin (clinician or patient unawareness of real risks but unclear benefits)

-- one issue is the availability of aspirin over-the-counter, which allows for unrestrained access to aspirin (which certainly has been around since time immemorial 😊  (actually it has been used >3500 years as a pain killer and antipyretic by Sumerians and Egyptians)

    -- for whatever reason (and likely drug company financial benefit), many meds are available OTC. And aspirin’s close relatives NSAIDs are also OTC, and (my guess, informed by my practice) very few people know the extent that NSAIDs are associated with so many of the chronic conditions that they may have, such as heart failure, hypertension, kidney failure, GI bleeding, bronchospasm…): the adverse effects are much worse than many of the prescription meds. there was an Albanian study finding very low health literacy about the adverse effects of aspirin, but nothing i have found in the western literature

    -- one other issue as mentioned in prior blogs: aspirin/NSAIDs may have increased major bleeding likelihood in patients who have untreated asymptomatic H Pylori infections (which is the most common bacterial infection in the world, over 50% of the global population is infected, and studies have suggested the rate in the US is around 25-30% and occurs in people who have never travelled abroad; treatment for H Pylori decreases this major bleeding risk by 2/3): https://gmodestmedblogs.blogspot.com/2013/04/h-pylori-erdication-if-on-aspirin.html , https://gmodestmedblogs.blogspot.com/2023/06/h-pylori-common-in-us-veterans.html (the latter referring to many other relevant studies)

 

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

  

to get access to all of the blogs:

 

 go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

  -- click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

  -- or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them

 

if you would like to see the articles in this blog, please email me. 

 

please feel free to circulate this to others. also, if you send me their emails (gmodest@bidmc.harvard.edu), i can add them to the list