Effects of a gluten-reduced or gluten-free diet for the primary prevention of cardiovascular disease

Background

Cardiovascular diseases are disorders of the heart and blood vessels, including acute coronary and cerebrovascular events (such as myocardial infarction and stroke). Approximately 50% of disabilities resulting from cardiovascular disease are associated with a suboptimal diet. In particular, dietary gluten (a protein found in certain cereal grain) has been linked to a wide range of adverse health outcomes. For example, different gastrointestinal symptoms (such as malabsorption or diarrhoea) often occur in people affected by gluten-related diseases and the only effective treatment for these populations is an often life-long gluten-reduced or gluten-free diet. Gluten-reduced and gluten-free diets have also gained a strong popularity in the general population. The evidence on the benefits and harms of a gluten-reduced or gluten-free diet in the general population is, however, contradictory: While gluten avoidance may be associated with disease prevention, there are also concerns that a gluten-free or gluten-restricted diet may be suboptimal (due to a reduced consumption of whole-grain, a dietary key component).

Review question

We reviewed the available studies to determine the effects (i.e. health-related benefits and risks) of a gluten-reduced or gluten-free diet for the primary prevention of cardiovascular disease in the general population.

Study characteristics

The evidence is current to June 2021. We identified one randomised controlled trial (RCT) and three non-randomised studies of interventions (NRSIs) with an observational design. The RCT was conducted in Italy and included 60 healthy adults with a follow-up of six months. The NRSIs included more than 450,000 participants (either health professionals from the USA or adults (volunteers) from the general population from the UK) and reported a maximum follow-up of more than 25 years. In the NRSIs, the lowest median gluten intake was 2.6 g/day and the highest was 9.4 g/day (thereby, 1 g/day of gluten intake is equivalent to half a slice of white bread). The RCT compared a gluten-free with a usual diet.

Key results

Our results suggest that it is unclear whether gluten intake is associated with all-cause mortality. Furthermore, our findings suggest no association between gluten intake and both cardiovascular mortality and non-fatal myocardial infarction. A lower compared with a higher gluten intake may be associated with a slightly increased risk to develop type 2 diabetes — a major cardiovascular risk factor. These findings were independent of age, ethnicity, body mass index (BMI), family history of diabetes, smoking status, alcohol intake, physical activity, menopausal status and postmenopausal hormone use, oral contraceptive use, multivitamin use, total energy intake, and intakes of magnesium, folic acid and cereal fibre.

The results from the only published RCT suggested that it is unclear whether gluten intake affects systolic blood pressure. Moreover, in this RCT no differences between a gluten-free and usual diet were found for other cardiovascular risk factors including diastolic blood pressure, low-density lipoprotein levels and BMI.

No study reported data on adverse events, quality of life or on other health-relevant outcomes.

Certainty of evidence

The certainty of evidence in this review was low to very low. The studies had several methodological flaws. Taking into account that the majority of the participants from the NRSIs were health professionals, the applicability to the general population is also questionable. Given the limited findings from this review predominantly based on observational studies, no recommendations for practice can be made.

Authors' conclusions: 

Very low-certainty evidence suggested that it is unclear whether gluten intake is associated with all-cause mortality. Our findings also indicate that low-certainty evidence may show little or no association between gluten intake and cardiovascular mortality and non-fatal myocardial infarction. Low-certainty evidence suggested that a lower compared with a higher gluten intake may be associated with a slightly increased risk to develop type 2 diabetes — a major cardiovascular risk factor. For other cardiovascular risk factors it is unclear whether there is a difference between a gluten-free and normal diet. Given the limited findings from this review predominantly based on observational studies, no recommendations for practice can be made.

Read the full abstract...
Background: 

Cardiovascular diseases (CVD) are a major cause of disability and the leading cause of death worldwide. To reduce mortality and morbidity, prevention strategies such as following an optimal diet are crucial. In recent years, low-gluten and gluten-free diets have gained strong popularity in the general population. However, study results on the benefits of a gluten-reduced or gluten-free diet are conflicting, and it is unclear whether a gluten-reduced diet has an effect on the primary prevention of CVD.

Objectives: 

To determine the effects of a gluten-reduced or gluten-free diet for the primary prevention of CVD in the general population.

Search strategy: 

We systematically searched CENTRAL, MEDLINE, Embase, CINAHL and Web of Science up to June 2021 without language restrictions or restrictions regarding publication status. Additionally, we searched ClinicalTrials.gov for ongoing or unpublished trials and checked reference lists of included studies as well as relevant systematic reviews for additional studies.

Selection criteria: 

We planned to include randomised controlled trials (RCTs) and non-randomised studies of interventions (NRSIs), such as prospective cohort studies, comparing a low-gluten or gluten-free diet or providing advice to decrease gluten consumption with no intervention, diet as usual, or a reference gluten-intake category. The population of interest comprised adults from the general population, including those at increased risk for CVD (primary prevention). We excluded cluster-RCTs, case-control studies, studies focusing on participants with a previous myocardial infarction and/or stroke, participants who have undergone a revascularisation procedure as well as participants with angina or angiographically-defined coronary heart disease, with a confirmed diagnosis of coeliac disease or with type 1 diabetes.

Data collection and analysis: 

Two review authors independently assessed eligibility of studies in a two-step procedure following Cochrane methods. Risk of bias (RoB) was assessed using the Cochrane risk of bias tool (RoB2) and the 'Risk Of Bias In Non-randomised Studies — of Interventions' (ROBINS-I) tool, and the certainty of evidence was rated using the GRADE approach.

Main results: 

One RCT and three NRSIs (with an observational design reporting data on four cohorts: Health Professionals Follow-up Study (HPFS), Nurses' Health Study (NHS-I), NHS-II, UK Biobank) met the inclusion criteria. The RCT was conducted in Italy (60 participants, mean age 41 ± 12.1 years), two NRSIs (three cohorts, HPFS, NHS-I, NHS II) were conducted across the USA (269,282 health professionals aged 24 to 75 years) and one NRSI (Biobank cohort) was conducted across the UK (159,265 participants aged 49 to 62 years). Two NRSIs reported that the lowest gluten intake ranged between 0.0 g/day and 3.4 g/day and the highest gluten intake between 6.2 g/day and 38.4 g/day. The NRSI reporting data from the UK Biobank referred to a median gluten intake of 8.5 g/day with an interquartile range from 5.1 g/day to 12.4 g/day without providing low- and high-intake categories.

Cardiovascular mortality

From a total of 269,282 participants, 3364 (1.3%) died due to cardiovascular events during 26 years of follow-up. Low-certainty evidence may show no association between gluten intake and cardiovascular mortality (adjusted hazard ratio (HR) for low- versus high-gluten intake 1.00, 95% confidence interval (CI) 0.95 to 1.06; 2 NRSIs (3 cohorts)).

All-cause mortality

From a total of 159,265 participants, 6259 (3.9%) died during 11.1 years of follow-up. Very low-certainty evidence suggested that it is unclear whether gluten intake is associated with all-cause mortality (adjusted HR for low vs high gluten intake 1.00, 95% CI 0.99 to 1.01; 1 NRSI (1 cohort)).

Myocardial infarction 

From a total of 110,017 participants, 4243 (3.9%) participants developed non-fatal myocardial infarction within 26 years. Low-certainty evidence suggested that gluten intake may not be associated with the development of non-fatal myocardial infarction (adjusted HR for low versus high gluten intake 0.99, 95% CI 0.89 to 1.10; 1 NRSI (2 cohorts)). Lowering gluten intake by 5 g/day also showed no association on the primary prevention of non-fatal and fatal myocardial infarction (composite endpoint) in linear dose-response meta-analyses (adjusted HR 1.02, 95% CI 0.98 to 1.06; 1 NRSI (2 cohorts)).

Coronary risk factors 

Type 2 diabetes

From a total of 202,114 participants, 15,947 (8.0%) developed type 2 diabetes after a follow-up between 22 and 28 years. There was low-certainty evidence that a lower compared with a higher gluten intake may be associated with a slightly increased risk to develop type 2 diabetes (adjusted HR 1.14, 95% CI 1.07 to 1.22; 1 NRSI (3 cohorts)). Furthermore, lowering gluten intake by 5 g/day may be associated with a slightly increased risk to develop type 2 diabetes in linear dose-response meta-analyses (adjusted HR 1.12, 95% CI 1.08 to 1.16; 1 NRSI (3 cohorts)).

Blood pressure, low-density lipoprotein level, body mass index (BMI)

After six months of follow-up, very low-certainty evidence suggested that it is unclear whether gluten intake affects systolic blood pressure (mean difference (MD) -6.9, 95% CI -17.1 to 3.3 mmHg). There was also no difference between the interventions for diastolic blood pressure (MD -0.8, 95% CI -5.9 to 4.3 mmHg), low-density lipoprotein levels (MD -0.1, 95% CI -0.5 to 0.3 mmol/L) and BMI (MD -0.1, 95% CI -3.3 to 3.1 kg/m²). 

No study reported data on adverse events or on other outcomes. Funding sources did not appear to have distorted the results in any of the studies.