Risk Score Reveals Nonvascular Factors Linked to GI Bleed

Liam Davenport

September 05, 2019

PARIS — Nonvascular risk factors such as frailty and previous conditions may indicate how likely a patient is to have major gastrointestinal (GI) bleeding when they take low-dose aspirin for the primary prevention of cardiovascular events, say UK scientists who developed a risk score from their findings.

Imen Hammami, PhD, Nuffield Department of Population Health, University of Oxford, and colleagues used data from the UK Biobank to examine nonvascular risk factors associated with major GI bleeds over 6 years.

Analysing Risk Factors

Their results, presented as a poster at the ESC Congress 2019, showed that, while events occurred at a rate of less than 0.1% per year, they were nevertheless able to identify a range of potential factors.

The resulting score, which included hand grip strength and prior respiratory haemorrhage, among other factors, showed that individuals in the highest risk category had a more than six-fold increased risk of major GI bleed.

The team says that the risk score "is applicable in other contexts" and will be tested using the ASCEND randomised trial of daily aspirin "to provide randomised evidence of the net benefit of aspirin in different risk categories".

They note, however, that their results also showed that "vascular risk factors predicted risk of GI bleeds, and nonvascular risk factors for GI bleed predicted risk of major vascular events, so the net benefit from aspirin may vary only modestly across categories by vascular and bleed risks".

Personal Risk 

Dr Kurt Huber, director of the Department of Cardiology and Intensive Care Medicine, Wilhelminen Hospital, Vienna, Austria, commented to Medscape News UK that, while the recorded event rate in the study was "very low", the findings underline that the primary prevention of vascular events should be personalised.

He said it is known that not all patients need both aspirin and an anti-platelet drug to avoid future ischaemic events.

"There may be some patients with several risk factors at the same time – hypertension, increased lipids, diabetes – where we might think about [that form of] primary prevention but it’s not something to give to every single patient."

Dr Huber added that "it’s an individual decision" that should take into account not only gastrointestinal but also intracerebral bleeding, as while this occurs in a "very low percentage" of patients, it’s "a huge problem when it happens".

Aspirin Risks and Benefits

The impetus for the research was that, while low dose aspirin prevents cardiovascular events, it also increases the risk of major, particularly GI, bleeds.

The team notes that the absolute benefit of aspirin increases with vascular risk, as does the bleeding risk, but it is less clear which nonvascular risk factors are linked to major GI bleeds.

To investigate further and determine whether individuals at low risk of bleed relative to their vascular risk could be identified, the team looked at data from the UK Biobank prospective study of 500,000 individuals aged 40–69 years.

They used self-reported diagnosis and electronic health records to ascertain history of cardiovascular disease, bleeds and other medical conditions, and gathered data on incident events over more than 8 years of post-recruitment follow-up.

The team used factors such as age, sex, smoking status, body mass index, systolic and diastolic blood pressure, diabetes, and antihypertensive drugs to stratify participants in terms of their baseline risk of a major vascular event.

This was defined as non-fatal myocardial infarction, non-fatal stroke or transient ischemic attack, arterial revascularisation or vascular death.

Study Findings

Among the individuals with no history of major vascular event, major bleed or GI ulcer in the 6 months prior to recruitment, 3522 individuals experienced a major GI bleed during follow-up, at a rate of 0.09% per year.

Among these, 2489 individuals had an upper GI bleed and 1235 a lower GI bleed.

In addition, 1404 participants had an intracranial bleed and 25,728 people had a minor bleed.

As expected, bleed rates increased with increasing risk for major vascular events. In addition, both antiplatelet use and anticoagulant use were associated with risks of GI bleed.

To identify nonvascular risk factors for major GI bleed, the team used stepwise regression analysis, with adjustment for established vascular risk factors and aspirin and warfarin use, from approximately 300 baseline and prior disease factors.

General factors associated with major GI bleed included:

  • ICD10-based frailty risk score, and frailty markers

  • Major and minor bleeds

  • GI diseases

  • Haematological factors

  • Cancers

  • Serum and urine markers for liver and kidney diseases

From this, the team derived a nonvascular risk score that included prior GI bleed, alcohol-related GI disease, and respiratory haemorrhage, as well as low forced expiratory volume in one second, low handgrip strength and low haemoglobin levels.

They found that, compared with individuals with the lowest quintile of major GI bleed risk, those in the highest quintile had a hazard ratio for major GI bleed of 6.57.

The risk score also reasonably discriminated between individuals based vascular event risk, with 38% of those in the highest quintile for major GI bleed risk having a major vascular event risk of ≥1% a year and 15% having a lower vascular risk.

Dr Huber said that the factors in the risk score were "not really surprising", as frailty and comorbidities have been shown in other studies to be linked to event risk, but that to have such a score is "a good idea".

That it also gave an indication of the risk of ischaemic events underlined to Dr Huber why, when it comes to primary prevention, "we should still be very reluctant and really think about which type of patient needs an anti-platelet drug or an oral anticoagulant".

The research was supported by grants from the UK Medical Research Council and the British Heart Foundation.

No conflicts of interest declared.

ESC Congress 2019: Abstract P656 . Presented August 31.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.