Mediterranean Diet: PREDIMED and Lyon Trial Evidence and an Indian Adaptation

The Mediterranean diet is a way of eating, not a branded plan: plenty of vegetables, fruit, pulses, whole grains and nuts, fish more often than red meat, and an unsaturated oil as the main fat. Its best known test is PREDIMED, a Spanish trial of 7,447 adults at high cardiovascular risk, in which the two Mediterranean groups had hazard ratios of 0.69 and 0.72 for heart attack, stroke or cardiovascular death against a reduced-fat diet. This entry sits in NNWA's handbook of diets beside the DASH diet.

What is the Mediterranean diet?

It is the food pattern of Crete, much of the rest of Greece and southern Italy in the early 1960s, as set out in a 1995 paper in the American Journal of Clinical Nutrition: abundant plant foods, fresh fruit as the usual dessert, olive oil as the principal fat, fish and poultry in low to moderate amounts, zero to four eggs a week and little red meat. Saturated fat was at or below 7 to 8 per cent of energy, while total fat ranged from under 25 to over 35 per cent.

Researchers need something tighter. The 2019 Cochrane review required a high ratio of monounsaturated to saturated fat, from olive oil or foods such as tree nuts, and a high intake of plant foods including fruit, vegetables and legumes. PREDIMED scored its participants with a 14-question screener, one point a question. The main thresholds are below.

What did PREDIMED and the Lyon trial find?

PREDIMED recruited 7,447 people aged 55 to 80, 57 per cent of them women, who had no cardiovascular disease but were at high risk of it. They were assigned to a Mediterranean diet with free extra-virgin olive oil, a Mediterranean diet with free mixed nuts, or advice to reduce dietary fat. After a median of 4.8 years, a heart attack, stroke or cardiovascular death had occurred in 3.8 per cent of the olive oil group, 3.4 per cent of the nut group and 4.4 per cent of the control group.

The Lyon Diet Heart Study was a secondary prevention trial: its 605 patients had already had a first heart attack. Those advised to follow a Mediterranean-type diet, and given a canola margarine to use, were compared with usual care. Over a mean of 46 months there were 14 cardiac deaths or non-fatal heart attacks in the Mediterranean group against 44 in the comparison group.

How strong is the evidence?

Better than for most diets, and still less certain than its reputation suggests. The 2019 Cochrane review covered 30 randomised trials with 12,461 participants and concluded that there is still some uncertainty about the effect of a Mediterranean-style diet on clinical events and on risk factors. It rated the evidence for modest improvements in risk factors in primary prevention as low or moderate in quality, and said there is a paucity of evidence for secondary prevention.

PREDIMED carries a history. Its 2013 report was withdrawn after the authors found that some participants had been assigned without proper randomisation; the 2018 paper is the corrected analysis, and results held after 1,588 such participants were left out. Cochrane read the trial as low-quality evidence of little or no effect on cardiovascular or total mortality, and moderate-quality evidence of fewer strokes, with a hazard ratio of 0.60.

Two further limits matter to an Indian reader. All three trials ran in Spain or France, and none read for this entry tested mustard or groundnut oil against olive oil. An Indian trial, the Indo-Mediterranean Diet Heart Study of 1,000 patients, appeared in the Lancet in 2002, but the journal issued an expression of concern about it in 2005, so this entry does not rely on its results.

Which Indian oil can stand in for olive oil?

On fat composition, groundnut oil is the closest everyday Indian match and mustard oil a reasonable second, but no trial has shown that either gives the PREDIMED result. The Indian Food Composition Tables 2017 do not list olive oil; they do give the fatty acid make-up of Indian oils. Mustard oil has the least saturated fat and carries alpha-linolenic acid, the plant omega-3, at 11.64 per cent. The two are set side by side on the mustard oil and groundnut oil comparison.

One detail keeps the comparison honest. In groundnut oil the whole 53.89 per cent of monounsaturated fat is oleic acid, the fatty acid olive oil is known for. In mustard oil only 10.18 per cent is oleic; 51.30 per cent is erucic acid, a different monounsaturated fat. ICMR-NIN's advice is to use different varieties of cooking oil, and the handbook's entry on omega-3 fatty acids explains why alpha-linolenic acid is worth having.

Quantity is the second difference. The screener gave a point for four tablespoons of olive oil or more a day. ICMR-NIN allows 27 to 30 g of visible fat on a 2,000 kcal day, within a range of 20 to 50 g that it equates to 4 to 10 teaspoons, and asks that ghee or butter be kept to one or two teaspoons. An Indian adaptation changes the kind of fat and keeps to the Indian amount.

An Indian day on the Mediterranean pattern

The day below is an illustration, not a prescription. It follows ICMR-NIN's My Plate for the Day for about 2,000 kcal: 250 g of cereals and millets, 85 g of pulses, 400 g of vegetables, 100 g of fruit, 300 ml of milk or curd, 35 g of nuts and seeds and 27 g of oil. The calorie calculator gives a starting figure.

Two notes. The screener's last point is for sofrito, a sauce of tomato with onion, leek or garlic simmered in olive oil; treating an Indian onion, tomato and garlic masala as its counterpart is this page's reading, not a trial finding. And fish is optional: My Plate says eggs, fish or meat can replace part of the pulses. The fish and seafood tables list Indian species, and the heart-healthy recipes give dishes to build on.

A worked example: scoring one week against the pattern

This is an illustrative case, not a real person. Take a 52-year-old schoolteacher in Kolkata who eats fish, cooks in mustard oil and has been told by her doctor that her cholesterol needs watching. She scores her usual week against the screener and the Indian guideline together; her follow-up test and any decision about medicine stay with the doctor.

Cost of the pattern in an Indian kitchen

No rupee figure is given here, because prices change by city and season and none was read from a primary source; the Cochrane review found that no trial reported on costs. What can be said is that the pattern does not depend on its expensive items. Olive oil is not needed for an adaptation, the screener counts peanuts as nuts, and a few walnuts from the nut table can be an occasional addition. Pulses and vegetables are already the base of an Indian meal.

Who does the Mediterranean diet suit?

It suits adults who want a long-term pattern for heart health and who can cook most meals at home. The trial evidence is for people aged 55 to 80 at high cardiovascular risk and for patients who already have heart disease, under medical care. That is a large group in India: the ICMR-INDIAB study of adults aged 20 and over put dyslipidaemia at 81.2 per cent, abdominal obesity at 39.5 per cent and diabetes at 11.4 per cent.

It fits Indian habits unevenly. NFHS-5 found that about half of women and men aged 15 to 49 eat pulses and dark green leafy vegetables daily, but only 12.5 per cent of women eat fruit daily and 34.4 per cent never eat fish. Vegetarians, 29 per cent of women and 17 per cent of men in that survey, can follow the pattern with pulses and nuts in place of fish.

Who should see a doctor before starting?

Anyone on a blood thinner, anyone with kidney disease or diabetes, and anyone with a nut or fish allergy. The US National Library of Medicine's guidance on warfarin says to eat consistent amounts of vitamin K foods from week to week, not to eat large amounts of leafy green vegetables or certain vegetable oils, and to speak to the doctor before any change in diet. Two leafy sabzis a day is such a change.

People with kidney disease may have limits on pulses, nuts and fruit; see the guide to a chronic kidney disease diet in India and ask the treating team. People with diabetes should agree fruit and grain portions with their doctor. The pattern is eaten beside prescribed treatment: it is not a reason to stop or reduce a statin or any other medicine. Chest pain, breathlessness or sudden weakness needs emergency care.

Is wine part of the diet?

It is part of the description, and it is not advice. The 1995 paper lists wine in low to moderate amounts with meals, and the PREDIMED screener gave a point for seven or more glasses a week. WHO's 2023 statement is that no level of alcohol consumption is safe for health and that alcohol causes at least seven types of cancer. Nobody should start drinking to follow this diet.

Common myths and mistakes

The opposite mistake is to add the famous foods to an unchanged diet: oil poured over a meal already cooked in oil, or a bowl of nuts after dinner. The pattern is a set of swaps, with red meat, butter, sweet drinks and bakery sweets going down as plant foods go up.

Learning this with NNWA

NNWA, a private skill academy in Kolkata, teaches online in English and Hindi. Its Mediterranean Diet Nutrition course and its wider programmes prepare people to coach healthy clients on patterns such as this one and to refer medical cases to a doctor. No NNWA course is a degree or a dietitian qualification.