Mediterranean, Low-Fat Diets Both Good for Health: Network Meta-analysis

Other options, like very low fat, Ornish, and Pritikin, didn’t offer the same reduction in death and nonfatal MI, researchers found.

Mediterranean, Low-Fat Diets Both Good for Health: Network Meta-analysis


(UPDATED) Both low-fat and Mediterranean diets offer health benefits and are good at preventing death and nonfatal MI, a network meta-analysis suggests.

However, others like the very-low-fat, Ornish, and Pritikin diets have no significant impact on either all-cause mortality or various cardiovascular outcomes, Giorgio Karam (University of Manitoba, Winnipeg, Canada), and colleagues report in the BMJ.

“Although dietary guidelines have suggested that a number of dietary programs might reduce the risk of major cardiovascular events, they have typically relied on surrogate outcomes, or low- or very-low-certainty evidence from nonrandomized study designs,” they note.

And while there have been various meta-analyses with two-group comparisons, approaching the array of options through a network meta-analysis allows for “the use of direct (head-to-head active interventions) and indirect evidence (intervention vs nonactive control) for the comparison of interventions that have not been directly compared, which can yield more precise summary estimates,” according to the researchers.

Karam, in an email to TCTMD, cautioned that while their methods allow them to describe their “certainty in the presence or absence of an effect,” they don’t capture “our certainty in the precise magnitude of effect.”

However, “most patients would find any reduction in events like mortality meaningful,” he said. And even with a degree of imprecision, “our risk differences can still be used by physicians to counsel patients about the general range of expected benefit from starting a structured dietary program, taking into account the point estimate and 95% CI.”

As for how to choose between the Mediterranean and low-fat routes, senior author Bradley C. Johnston, MD (Texas A&M University, College Station), pointed out in an email that the risk reductions they found tended to be larger with Mediterranean, but he also said the evidence is too limited to be definitive.

“Evidence-based clinical practice is not only just about evidence, but also incorporating patient values, preferences, cost, and potential inconvenience of available therapies,” Johnston told TCTMD, noting that all of these could play a role in decisions about diet.

Safi U. Khan, MD (West Virginia University, Morgantown, WV), who didn’t take part in the study, agreed that doing a network meta-analysis “provides valuable insights” that can be applied to the question of which dietary program is best.

What’s more, he said in an email, “the authors provide absolute estimates and certainty of estimates for patients at intermediate and high risk using GRADE methods, which can help clinicians and policy makers make informed decisions about the most effective dietary interventions for the specific patient population.”

He, like the authors in their paper, pointed out that the analysis didn’t include data from the 2022 CORDIOPREV trial favoring a Mediterranean diet over low fat. Adding that trial into the mix “might have shifted the results more definitively in favor of the Mediterranean diet,” Khan suggested, noting that without actually reanalyzing the data this idea is only speculative.

Evidence-based clinical practice is not only just about evidence, but also incorporating patient values, preferences, cost, and potential inconvenience of available therapies. Bradley C. Johnston

Karam and colleagues identified 40 eligible trials with a total of 35,548 participants that looked at seven different dietary programs:

  • 18 studies of low fat (total fat 20% to 30% of caloric intake and saturated fat < 10% of caloric intake)
  • 12 studies of Mediterranean (emphasis on fish, fruits, vegetables, and monounsaturated fats)
  • Six studies of very low fat (fat intake 10% to 20% of caloric intake)
  • Four studies of modified fat (increase in polyunsaturated-to-saturated fat ratio)
  • Three studies of combined low fat/low sodium (like low fat but with sodium < 2.4 g/day)
  • Three studies of Ornish (fat < 10% of caloric intake and primarily plant-based)
  • One study of Pritikin (carbohydrates 70% to 75%; protein 15% to 20%, and fat 5% to 10% of caloric intake plus fiber intake of 40 to 45 g/1,000 kcal)

These diets were compared against minimal intervention, which consisted of things like usual diet or no/minimal advice, patients being referred to their physician, and programs unrelated to diet.

The Mediterranean diet was linked, by moderate-certainty evidence, to less all-cause mortality, CV mortality, stroke, and nonfatal MI compared with minimal dietary intervention in people at intermediate risk, whose predicted rate of CV events at baseline was 5% to 10% over 5 years.

Mediterranean Diet vs Minimal Intervention: Intermediate-Risk Patients


OR (95% CI)

Fewer Cases per 1,000 People Over 5 Years

All-Cause Death

0.72 (0.56-0.92)


CV Mortality

0.55 (0.39-0.78)



0.65 (0.46-0.93)


Nonfatal MI

0.48 (0.36-0.65)


Also based on moderate-certainty evidence, low-fat diets were better than minimal intervention at reducing all-cause death and nonfatal MI. They also were linked to fewer unplanned CV interventions, but with low-certainty evidence.

Low-Fat Diet vs Minimal Intervention: Intermediate-Risk Patients


OR (95% CI)

Fewer Cases per 1,000 People Over 5 Years

All-Cause Death

0.84 (0.74-0.95)


Nonfatal MI

0.77 (0.61-0.96)


Unplanned CV Interventions

0.57 (0.35-0.93)


In patients at high baseline risk, for whom the projected rate of CV events was 20% to 30% over the next 5 years, the benefits were even greater. The Mediterranean diet was linked to 36 fewer all-cause deaths, 39 fewer CV deaths, 16 fewer strokes, and 42 nonfatal MIs per 1,000 people over 5 years. The low-fat diet was linked to 20 fewer all-cause deaths, 17 fewer CV deaths, and 18 fewer nonfatal MIs.

Yet it’s impossible to rule out the potential for benefit with the other five diets.

Karam said that for some of those they studied, like Ornish, their certainty about a lack of effect was limited by small sample sizes, which produced wide confidence intervals. “Even for dietary programs where our certainty in the absence of effect was not limited by imprecision, we had concerns that poor adherence to the intervention may have been a contributing factor,” such as for the combined low fat/sodium diet, he added.

Going forward, more research would be welcome to clarify which aspects of interventions—diets with or without physical activity, behavioral support, smoking cessation, or drug therapy—“are most essential,” said Karam. Additionally, “randomized comparisons would be ideal,” he noted. Examples include the PREDIMED-Plus trial, which has released 1-year results, as well as MeMeMe and the AUSMED Heart Trial.

Both US and European prevention guidelines, based on the results of the PREDIMED trial and other research, endorse the Mediterranean diet. 

Khan observed that the studies included in the current meta-analysis involved not only diet alone but also sometimes diet plus other interventions like exercise and psychosocial support, thus “emphasizing the importance of a holistic, healthy lifestyle approach.”

Overall, the results “validate the narrative that adopting cardioprotective Mediterranean or low-fat diets, regular exercise, and lifestyle modifications can effectively reduce cardiovascular risk,” he concluded. “Clinicians should consider patients' preferences, lifestyles, and cultural backgrounds when recommending a sustainable diet and overall health plan.”

Caitlin E. Cox is News Editor of TCTMD and Associate Director, Editorial Content at the Cardiovascular Research Foundation. She produces the…

Read Full Bio
  • Karam and Khan report no relevant conflicts of interest.
  • Johnston received a grant from Texas A&M AgriLife Research to fund investigator-initiated research related to saturated and polyunsaturated fats. The grant was from Texas A&M AgriLife institutional funds from interest and investment earnings, not a sponsoring organization, industry, or company.