Practical dietary interventions to prevent cardiovascular disease suitable for implementation in primary care: an ADAPTE-guided systematic review of international clinical guidelines

Abstract

Purpose

Cardiovascular diseases (CVD) are the leading cause of death globally. The current model of care for high-income countries involves preventive medication and highly trained healthcare professionals, which is expensive and not transposable to low-income countries. An innovative, effective approach adapted to limited human, technical, and financial resources is required. Measures to reduce CVD risk factors, including diet, are proven to be effective. The survey “Scaling-up Packages of Interventions for Cardiovascular disease prevention in selected sites in Europe and Sub-Saharan Africa” aims to develop non-pharmacological cardiovascular prevention and control programs in primary care and community settings in high, middle, and low-income countries. This review aims to identify the existing, validated dietary interventions for primary CVD prevention from national and international clinical guidelines that can be implemented in primary care and communities.

Methods

A systematic review of CVD prevention guidelines was conducted between September 2017 and March 2023 using the Turning Research Into Practice medical database, the Guidelines International Network, and a purposive search. The ADAPTE procedure was followed. Two researchers independently conducted the searches and appraisals. Guidelines published after 01/01/2012 addressing non-pharmacological, dietary interventions for primary CVD prevention or CVD risk factor management, in the adult general population in primary care or in community settings were included and appraised using the Appraisal of Guidelines Research and Evaluation II score. Individual dietary recommendations and the studies supporting them were extracted. Then supporting data about each specific dietary intervention were extracted into a matrix.

Results

In total, 1375 guidelines were identified, of which 39 were included. From these, 383 recommendations, covering 10 CVD prevention themes were identified. From these recommendations, 165 studies for effective dietary interventions for CVD prevention were found. Among these, the DASH diet was the most effective on multiple CVD risk factors. Combining diet with other interventions such as exercise and smoking cessation increased efficacy. No guidelines provided detailed implementation strategies.

Conclusion

The DASH diet combined with other interventions was the most effective on an individual basis. However, expansion in the wider population seems difficult, without government support to implement regulations such as reducing salt content in processed food.

Trial Registration

Clinical Trials NCT03886064

Plain English summary

Heart disease is the leading cause of death around the world. Strategies to prevent heart disease in high-income countries rely on medications and the skills of highly trained healthcare professionals. However, this is expensive and unsuitable for low-income countries. Consequently, an innovative, effective approach, which can be adapted to countries with limited human, technical and financial resources is needed. A program called SPICES was developed to identify strategies other than medication to prevent and control heart disease. This program reviewed the evidence for smoking cessation, physical activity, and dietary strategies, which may be useful to prevent heart disease in communities with limited resources.

In this review, the investigators searched online databases to find clinical guidelines that recommended dietary strategies to manage heart disease worldwide. The information found from this search revealed that the DASH diet, inspired by the Mediterranean diet, helps with weight loss, and improves blood pressure and cholesterol levels making it the most effective diet for preventing heart disease. It is even more effective if it is combined with other strategies such as exercise, stopping smoking or reducing the amount of alcohol consumed. However, this works well for individuals but is difficult to expand to the wider population. Therefore, government support is needed to implement regulations such as reducing salt content in processed food.

Introduction

During the twenty-first Century, cardiovascular diseases (CVD) have become the leading global cause of death, resulting in 17.9 million deaths in 2019 [1]. In Europe, CVD costs are estimated at 169 billion Euros annually [2]. Furthermore, CVD death rates are higher in the lower socioeconomic levels between and within countries with three quarters of CVD deaths occurring in low-income countries [3].

The current model of care for high-income countries involves preventive medication and highly trained healthcare professionals but is becoming increasingly difficult to maintain due to high costs and is not transposable to low-income countries. An innovative, effective approach adapted to limited human, technical, and financial resources is required. Inspired by the progress made in HIV and AIDS treatment in Sub-Saharan Africa (SSA), the World Health Organization (WHO) created the ICCC Framework (Innovative Care for Chronic Conditions). This framework redirects the current health policies from being population-centered to patient-centered, involving patients, families, and communities [3].

CVD occurrence increases predictably with accumulating CVD risk factors. Measures to reduce modifiable CVD risk factors, including diet, are proven to be effective. For example, in the general population, it is estimated that reducing cholesterol by 10% could decrease CVD mortality by 20% [4] and adopting a Mediterranean diet could reduce cardiovascular event occurrence by 30% [5]. Furthermore, CVD prevention interventions such as policies promoting healthy eating, smoking cessation, and physical activity, are cost-effective on international and national levels. It has been estimated that their costs would not exceed 4% of current health expenditure in high-income countries and 1–2% in low-income countries [6]. Also, behavioral changes are cost-effective on an individual level [7].

Consequently, the “Scaling-up Packages of Interventions for Cardiovascular disease prevention in selected sites in Europe and Sub-Saharan Africa” (SPICES) study was developed to evaluate and implement a comprehensive non-pharmacological cardiovascular prevention and control program in primary care and community settings in high, middle, and low-income countries [8]. The study involved systematically reviewing systematic clinical practice guidelines for smoking cessation, physical activities, and diet to identify best practice recommendations for reducing CVD. The smoking cessation and physical activity reviews have been previously published [9].

This review aims to identify the existing, validated dietary interventions for primary cardiovascular prevention from national and international clinical practice guidelines including European Union Countries, the UK, and Sub–Saharan Africa that can be implemented in primary care and communities.

Material and methods

Information sources and search

A systematic review of CVD prevention guidelines was conducted at the beginning of the SPICES study between September 2017 and January 2018 using the TRIP (Turning Research Into Practice) medical database and the International Guidelines Library of the Guidelines International Network (G-I-N). Subsequently, an update was performed in March 2023 prior to publication to ensure all recent guidelines were included. Following the ADAPTE procedure, the PIPOH tool (Population, Intervention, Professional/Patient, Outcome, Healthcare setting) was used to define the database search queries [10]. Population was defined as a primary care general population, free from cardiovascular disease. The Interventions were those focusing on cardiovascular risk factors including diabetes, hypertension, smoking, sedentary lifestyle, unhealthy diet, excess weight, or obesity. Professionals/Patients were any healthcare professional working in primary care or lay people. The Outcomes were reduced morbidity or mortality. The Healthcare setting and context was primary care. Search queries were “cardiovascular disease prevention” and “cardiovasc* prevention OR risk* OR risico* OR risque*”.

Then, a purposive search for every national clinical guideline used in the SPICES countries was performed. Guidelines included the Haute Autorité de Santé for France, the Tijdschrift Huisarts and the Nederland Huisartsen Genootschap for Belgium, the National Institute for Health and Care Excellence (NICE) for the United Kingdom, the European Society of Cardiology for European Countries for South Africa and the WHO for Uganda.

The review was reported following the 2020 PRISMA reporting guidelines.

Eligibility criteria

Guidelines were included if they addressed non-pharmacological, dietary interventions for primary CVD prevention or CVD risk factor management, in the adult general population in primary care or in community settings. At least one patient outcome measure used for CVD risk assessment, such as mortality and morbidity, had to be reported in the guidelines. They had to be published after 01/01/2012 and be the latest version for revised guidelines. Guidelines could be written in English, French or Dutch.

Guidelines were excluded if they focused solely on specific populations such as elderly people, infants, children, pregnant women, or people with cancer. Guidelines focusing on secondary or tertiary prevention or only addressing cardiovascular risk assessment, pharmacological or surgical interventions, or specific conditions (such as type 1 diabetes, familial hypercholesterolemia) were excluded as were guidelines published before 2012 as these were considered out-of-date. Guidelines with no free full text availability were excluded as the consortium felt that recommendations to healthcare professionals or stakeholders should be freely accessible.

Searches were independently conducted by two researchers with a merging of results at each step of the review. Discrepancies were resolved by the two researchers and the study scientific committee.

Guideline, recommendation, and study selection

Two data collection phases were performed. During the first phase, the two researchers independently selected the guidelines. In the second phase, guidelines were evaluated according to the “Appraisal of Guidelines Research and Evaluation II” (AGREE II) tool integrated in the ADAPTE procedure [10]. The AGREE II tool consists of 23 items arranged into six domains: scope and purpose, stakeholder involvement, rigor of development, clarity of presentation, applicability, and editorial independence. The two researchers independently scored each guideline by domain. An overall assessment was then performed and ranged from 1 (strongly disagree) to 7 (strongly agree). This score was independent from the other domain scores.

A scientific committee consisting of three members was created to select guidelines based on the overall assessment scores. If the overall assessment scores from both researchers were equal to or higher than 5 for a selected guideline, it was included in the review. The difference between the overall assessment scores could not exceed one. If the difference was greater than one, an overall assessment score was found by consensus between the two researchers and the scientific committee.

Once the guidelines were selected, they were inspected for specific recommendations for effective, dietary interventions for primary CVD prevention. Recommendations with an A or B level of evidence, or Class I or Strong and/or 1 +  + ,1 + , 2 +  + , 2 + for NICE grading (regardless of level of evidence) were included. If an effective dietary intervention for primary CVD prevention with a pragmatic implementation strategy was supported by a study, this study, regardless of date, was considered for inclusion in a final matrix. Study exclusion criteria included study population under 50, gender specific study, and the absence of a control group for individual interventions.

As the recommendations in the different guidelines frequently cited the same studies, a duplicate elimination process was performed. The complete selection process of guidelines, recommendations and studies is described in Fig. 1.

Fig. 1

Selection process of guidelines, recommendations, and studies

Data extraction and synthesis

For each included guideline, publication year or year of latest update, country, developing organization, language and title were extracted. From each included guideline a list of all dietary recommendations was created to provide an overview of relevant content. Recommendation themes were created inductively by grouping similar topic recommendations. The strength of recommendation, level of evidence, intervention description, outcomes, implementation strategies and evidence gaps were extracted from each recommendation if reported. The recommendations and studies they cited were then compiled into two matrices. After reading the full text, study title, implementation or intervention description, intervention frequency and duration, setting, material used, psychological model used where applicable, mass media use, and delivered intervention status were extracted into the matrices.

Results

A total of 39 primary CVD prevention guidelines were included (Fig. 2 and Table 1). Two guidelines were from the WHO and 37 were from high-income countries. None were by an organization from a middle- or low-income country.

Fig. 2

Selection of primary CVD prevention guidelines, recommendations and studies

Table 1 List of included guidelines addressing dietary interventions for primary CVD prevention

From the selected guidelines, 383 dietary recommendations were extracted. Recommendation ratings indicate comparability and quality. The Canadian Diabetes Association, Canadian Task Force on Preventive Health Care, National Institute for Health and Care Excellence (NICE), Department of Veterans Affairs and the Department of Defense (VA/DoD), World Health Organization (WHO) and the collaboration between the American Diabetes Association (ADA) and the European Endocrine Society used the independent GRADE rating system which is a current gold standard [11]. The other organizations used their own rating systems or other organization’s rating system. However, 11guidelines had no recommendation rating system, including six NICE guidelines. The body of evidence could be linked to the recommendations in 29 guidelines and could not in the other 10 (Table 1).

Among the 383 dietary recommendations extracted, ten major themes were identified (Table 2 and Table 3). Of these, only 17 recommendations recommended against current strategies (4.4%). Recommendations addressing soya, nuts, stanol esters and sterols were contradictory, with the same recommendation strength. Vitamin D, potassium, calcium, magnesium, vitamins A, B, C, E, and omega-3 fatty acids supplements were also controversial. For example, the Academy of Nutrition and Dietetics promoted magnesium, calcium, potassium, and vitamin D supplementation while the European Society of Cardiology (ESC) and the Scottish Intercollegiate Guideline Network (SIGN) stated that the evidence did not support supplementation [12,13,14]. Only one guideline recommended that no alcohol should be consumed when 13 recommended only consuming small amounts [15].

Table 2 Summary of diet and lifestyle recommendations

Table 3 Summary of diet and lifestyle implementation recommendations

Within the non-pharmacological recommendations, 1210 studies were identified, of which 164 were included in the diet study matrix (Table 4). Among these studies, 42 investigated dietary interventions targeting individuals, 84 investigated lifestyle interventions targeting individuals, and 38 involved communities in CVD prevention using dietary interventions. All but two of the 164 studies, used a CVD prevention surrogate endpoint [16, 17]. The mean publication dates were 1999 for dietary interventions, 2004 for lifestyle interventions and 1994 for studies involving communities. Only nine studies involving communities were randomized controlled trials. The selection criteria for this category was therefore lowered to cohort studies to include the other 29 studies.

Table 4 Matrix of dietary interventions

The Dietary Approaches to Stop Hypertension (DASH) diet was the only diet proven to be effective for the three main CVD risk factors: blood pressure, lipid profiles and weight loss [18,19,20,21].

Twenty-six studies showed effective dietary interventions to reduce blood pressure. Eleven were individual dietary interventions, 10 were lifestyle interventions and five were community engagement studies. The interventions tested in these studies included a weight loss diet in ten studies and sodium reduction in the other sixteen studies. The interventions were led by dieticians in 14 studies, and by community health workers in one study. The provider was not described for 11 interventions. There were two or more providers of dietary interventions in 13 studies.

Seventeen studies presented effective dietary interventions for improving lipid profiles. Diets tested included the American Heart Association (AHA) diet [22, 23], the Hospital Contribution Fund of Australia diet (HCF) [22], the National Cholesterol Education Program (NCEP) recommendations [24], low-fat diets [22, 25,26,27,28], low-carbohydrate diets [27, 29], Mediterranean diet [17, 26, 30], commercial diets [31], and the low-calorie DASH diet [19]. Interventions were educational programs and mainly involved dieticians, instructors, psychologists, and nurses.

Fifty-five studies revealed effective diets for weight-loss. Ten studies focused exclusively on diet. The greatest weight loss was achieved with a low-carbohydrate diet but every diet involving calorie restriction was effective [19]. A Mediterranean, DASH diet, or low-fat diets should be promoted depending on individual CVD risk factors and patient preferences [23]. Forty-four studies involved combined interventions with diet plus physical activities. Mean weight loss was similar between dietary and lifestyle interventions (around 5 kg). Interventions were mainly led by dieticians and nutritionists and ranged from dietary advice to structured educational programs.

Two studies showed that physicians providing brief advice effectively reduced alcohol consumption [32, 33].

Eighteen community studies evaluated changes in purchases in favor of healthier food choices such as fat-free options. Outcomes were diverse and included cash register data, gallons of consumed milk, and sales increases. However, no cardiovascular health benefits could be extrapolated from these results.

Thirty-three studies evaluated recommendations to participants to follow a healthier diet using questionnaires. However, none of the questionnaires used were validated, meaning it was difficult to extrapolate cardiovascular health benefits from these results. Multiple professionals led the interventions including teachers, local stakeholders, physicians, nutritionists, dieticians, nurses, and computer systems.

The PREDIMED study was the only study aimed at improving cardiovascular outcomes that used a direct CVD prevention endpoint defined as major cardiovascular events. It implemented an energy-unrestricted Mediterranean diet supplemented with extra-virgin olive oil or nuts in a high CVD risk population [17]. After a 5-year follow-up, the end-point event incidence rates were 8.1 per 1000 person-years for the Mediterranean diet with extra-virgin olive oil, 8.0 per 1000 person-years for the Mediterranean diet with nuts and 11.2 per 1000 person-years for the control (low-fat) diet. Those on the PREDIMED diet had a lower risk of major cardiovascular events over a 5-year period than those on the control (low-fat) diet, with a relative difference of 30%. Another study, involving participants from the Arthritis, Diet, and Activity Promotion Trial (ADAPT), showed that the mortality rate of participants randomized to an 18-month weight loss intervention was lower than controls (hazard ratio = 0.5, 95% confidence interval 0.3–1.0) seven years after the intervention [16].

Lifestyle interventions associated a dietary intervention with at least one other intervention and effectively reduced CVD morbidity. Eighty-five studies combined diet with between one and three different interventions including physical activity, lifestyle management, smoking cessation, alcohol reduction, and involved healthcare professionals. Twelve other studies involved communities and changes to workplace and public layouts. Sixty studies combined just diet and physical activity making this the most common combination. This was more effective on blood pressure [17], lipids [17], weight loss and body composition [17]. The effects of these lifestyle interventions were more sustained than diet alone [17]. Any physical activity type was effective if it was supervised by the research team. Professionals involved were diverse and included physicians, dieticians, nurses, and lay people.

The Finnish Diabetes Prevention study [34] and the Diabetes Prevention Program (DPP) [35] demonstrated that a structured lifestyle program including weight loss effectively reduced diabetes incidence. Six other studies included in the matrix studied the implementation of these programs. They were effective in multiple settings [36, 37] and had long-term efficacy [38]. Delivery of such programs using conference calls [39]or community health workers [40] was effective.

Discussion

To our knowledge, this is the first systematic review of international clinical guidelines following the ADAPTE process to identify effective dietary interventions for CVD prevention in primary care or community settings. A matrix of 383 effective dietary interventions to prevent CVD was created. The SPICES literature review was created to provide a matrix of effective, detailed interventions on diet, physical activity, and smoking regulation for CVD primary prevention to implement into vulnerable communities. For this purpose, the DASH diet appears to be the most effective in preventing CVD as it concurrently reduces blood pressure, improves the lipid profile, and leads to weight loss. This is a variation of the Mediterranean diet and provides indications about the amounts of each nutrient that should be consumed. It can be adapted to the cultural habits of every community and different CVD patient profiles. Combining diet with other interventions such as physical activity and smoking cessation increases efficacy. Physicians providing brief advice is shown to be an efficient strategy to address excessive alcohol consumption. Overall, guidelines did not provide detailed strategies to implement the recommendations in communities.

Guidelines had contradictory recommendations on supplements. They were all rated on their editorial independence following the AGREE II tool. External review of their content was performed before publishing. However, their conclusions were contradictory despite being based on the same body of evidence. External review should admonish institutional bias. However, external review within the same profession can lead to opinions that are distanced from the evidence.

The different dietary interventions were led by a variety of professionals including physicians, dieticians, health workers, nurses, and lay people. This is encouraging as it reveals that diverse professionals can successfully implement and lead these interventions, which is important for areas with poor access to healthcare professionals. Furthermore, training lay people to lead dietary interventions for CVD prevention could be a possible solution for public health authorities, particularly in low-income countries.

This review underlines the lack of recommendations and studies adapted to the low- and middle-income countries. In fact, only five studies in the final matrix were conducted in these countries [21, 41,42,43,44]. This meant stakeholders and researchers had to estimate the transferability of high-income country prevention strategies, especially for dietary interventions. However, the similar spread of junk food in these countries has already been demonstrated [45].

The matrix produced from this review enables any researcher or caregiver to choose a validated intervention to implement in primary care. However, some intervention implementation details, implied psychological models and intervention durations, are missing from the matrix as they were not described in the studies or were too old. This affects the reproducibility of these effective interventions, and the imprecisions could lead to failure. Furthermore, implementing the current, very general recommendations may miss the precise key required for interventions to be successful.

The selected studies mainly focused on individual changes rather than community approaches (126 studies versus 38 studies). The individual benefit of dietary interventions for primary CVD prevention was small in some studies, for example losing a few kilograms in weight or improving blood pressure by several millimeters of mercury. This therefore raises the question as to whether it is ethical to ask individuals to commit such efforts. Furthermore, during clinical trials, individuals find it easier to stick to a dietary intervention but in real-life situations, they are often unlikely to make long-term, significant dietary changes. Cheaper, more plentiful, higher-fat foods make following strict diets harder and also contribute to the increasing prevalence of CVD risk factors such as obesity, high blood pressure and increased lipid levels making primary CVD prevention an important public health issue [46]. Since the included studies presented interventions for individuals, it was difficult to extrapolate the results to the general population. However, some individual interventions such as salt restriction could be implemented in the general population for example by regulating salted food items. This could reduce salt intake and subsequently reduce blood pressure therefore reducing CVD mortality [47].

Strengths and limitations

One study strength was using international clinical guidelines as information sources with searches being conducted on specific guideline databases, TRIP and G-I-N. However, these databases were incomplete, meaning other databases and a purposive search were required to complete the selection process and find the missing guidelines. The selection process was rigorous due to the two double-blind phases and the involvement of a scientific committee. Information bias was controlled through evaluation triangulation. It was assumed that selection and publication biases were previously addressed by each guideline provider. Assessing each guideline with the AGREE II tool meant that the researchers expected to select guidelines with the most reliable bibliographies and intervention implementation information could be captured. Some authors suggest selecting guidelines just by scoring the rigor and development domain [48]. However, this misses the stakeholder involvement and applicability domains which are other areas of concern for implementation. This review had no specific strategy to assess risk of individual study biases. It was assumed that the guideline developers dealt with this. However, it should be limited by retaining only high-quality guidelines which is one of the reasons for using the ADAPTE process.

Searching for implementation information resulted in a deeper search into the guidelines. This produced an underestimated workload that ADAPTE was supposed to prevent. An evaluation cascade occurred, firstly of guidelines, then recommendations and finally cited studies. The diverse recommendation rating systems and the fact that links between recommendations and the bodies of evidence were sometimes missing meant selecting recommendations with a high level of evidence was impossible. Mean study publication dates were unexpectedly out of date. Guidelines should cite the most up to date studies and this review shows that this is not the case. It is unclear whether this is because guidelines do not regularly update their body of evidence or whether there is little recent research on dietary interventions. No common measure was found to present study efficacy since population, outcomes, and statistical processes were disparate. Even for highly recommended interventions, only trends can be estimated. Furthermore, recommendations are based on a body of evidence that is largely based on surrogate endpoints.

Conclusion

On an individual basis, this review indicates that implementing the DASH diet combined with any kind of physical activity, smoking reduction or cessation, and brief advice about reducing alcohol consumption effectively reduces CVD risk factors. However, this seems difficult to expand to the wider population without government support to implement regulations such as reducing salt content in processed food.