More fruit or vegetables did not line up with lower BMI in a 641-person Saudi survey. The online convenience sample was young, educated and unable to show which came first

Does eating fruit and vegetables more often show up as a lower body-mass index in a population where both obesity and low produce intake are common? Researchers distributed an online food-frequency questionnaire through social media and received responses from 641 adults living in Saudi Arabia.

Among the 635 people with usable self-reported height and weight, 43.5% were in the normal BMI range, 29.8% were overweight and 21.7% were in an obesity category. The sample was not a miniature Saudi population: 68.6% were younger than 35 and 78.6% had a bachelor's or postgraduate degree.

Fruit frequency did not differ clearly across BMI groups, with p=0.14. The median was actually highest in the overweight group at 50 times a month, compared with 47 in the normal-weight group, 34 in the underweight group and 40 in obesity class I. Vegetable frequency also missed statistical significance, at p=0.71; normal-weight participants reported a median 88 times a month and obesity groups roughly 70 to 80.

A cross-sectional questionnaire cannot tell whether diet influenced weight, weight influenced what people ate, or both reflected income, health, appetite and reporting habits. Frequency also says nothing about portion size, cooking method or the rest of the diet. Fruit and vegetables remain useful for fibre, micronutrients and dietary variety; this survey simply could not turn how often people reported them into a reliable explanation of BMI.