We were interested in the strong associations found by Wium-Andersen et al between ischaemic heart disease and stroke on the one hand and depression on the other.^{Reference Wium-Andersen, Wium-Andersen, Prescott, Overvad, Jørgensen and Osler1} They found no intervening variable that could explain the relationship. Diet was not investigated. Although studies of nutrition and diet can be complex and imprecise, there is evidence that a Mediterranean diet protects against both depression and cardiovascular disease.^{Reference Fresán, Bes-Rastrollo, Segovia-Siapco, Sanchez-Villegas, Lahortiga and de la Rosa2,Reference Salas-Salvadó, Becerra-Tomás, García-Gavilán, Bulló and Barrubés3} The authors had access only to total serum cholesterol levels, but non-high-density lipoprotein (non-HDL) cholesterol may have been relevant. Low HDL and high non-HDL-cholesterol can be risk factors for both depressions and cardiovascular disease.^{Reference Lehto, Niskanen, Tolmunen, Hintikka, Viinamäki and Heiskanen4,Reference Ridker, Rifai, Cook, Bradwin and Buring5} These parameters are heavily influenced by diet. The authors were also constrained by imprecise data on obesity, which was estimated in the study by body mass index (BMI). Obesity is not accurately measured by BMI. A person's percentage of body fat is a good measure of obesity, and this correlates both with depression and with cardiovascular disease, even among people with a BMI within the ‘normal’ range.^{Reference Chen, Arthur, Iyengar, Kamensky, Xue and Wassertheil-Smoller6} In our opinion, this study could not rule out the possibilities that diet and/or obesity are important intervening variables between ischaemic heart disease, stroke and depression.