At 3,825 metres in the Peruvian Andes, a better diet went with lower blood pressure only in people already treated for it

Most of what we know about diet and blood pressure comes from studies at sea level in wealthy countries, where the DASH and Mediterranean diets have been tested again and again. Life in the high Andes looks very different: cold, thin air, a food culture built on potatoes and grains, and patchy access to clinics and medicines. A team from Johns Hopkins University, Universidad Peruana Cayetano Heredia, Washington University in St. Louis and partners in Puno set out to measure what adults with high blood pressure there actually eat, and whether eating better goes with lower readings.

Their study, published on 28 October in PLOS Global Public Health, used baseline data from the ANDES trial, which tests a hypertension programme run by community health workers in Puno, a region of southern Peru at 3,825 metres above sea level. Between February and August 2024, trained local nutritionists visited 154 adults with hypertension at home (average age 66.5, 52% women) and took a detailed account of everything they had eaten and drunk the day before, weighing food in the house where they could. Each diet was scored with the Global Diet Quality Score, which runs from 0 to 49 and rewards 16 healthy food groups, such as fruit, leafy greens, legumes and whole grains, while taking points off for 9 groups such as refined grains, sweets, sugary drinks and white roots and tubers. Blood pressure was measured three times with a validated digital monitor.

Diets were poor almost across the board. The average score was 15.7 out of 49, and only two people reached the 23 points linked with low risk; 43% scored under 15, the high-risk band. Few people ate deep orange fruit, cruciferous vegetables or low-fat dairy, and 135 of the 154 ate white roots and tubers, mainly potatoes, averaging 262 grams a day among those who did. There was a bright side: only 3 people ate processed meat, 14 ate bought fried food and 32 had sugary drinks. Across the whole group, a higher score showed no clear link with blood pressure once age, wealth, food security, weight, activity, alcohol and blood sugar were taken into account. Among the 102 people who knew they had hypertension and were taking medicine, though, each step up in diet-score quartile went with systolic pressure about 4.1 mmHg lower, and in the third quartile it was 18.2 mmHg lower than in the lowest.

The authors are careful about what this means. Diet and blood pressure were measured once, at the same time, so the study cannot show that the better diet lowered pressure. People who are diagnosed and treated may also keep appointments, take their tablets and follow lifestyle advice more closely, which could explain both the better diet and the lower readings. A single day's recall can miss usual eating, salt was not measured, the subgroup results were exploratory and not adjusted for multiple testing, and the formal test for a difference between the treated and untreated groups was not significant. The score also counts every potato as unhealthy, however it is cooked, which may unfairly drag down a diet where boiled potatoes are the staple and fries are rare. The team suggests that nutrition advice delivered at home by community health workers, built around local foods and cooking, belongs in hypertension programmes like the one ANDES is testing.