Demonstration of similarities in geographic distribution of multiple sclerosis (MS) and component of diet would strengthen a consideration that diet plays a causal role in MS. Prevalence of MS in many countries was correlated with average daily per capita consumption of fats and oils, protein, and calories, including calories of animal origin. Of these dietary factors, only calories of animal origin and fats and oils correlated significantly with MS prevalence. When the latter two were combined (animal-fats), a significant correlation with MS prevalence of .70 resulted, suggesting that increased consumption of animal-fat may be associated with MS. Experimental and clinical observations relating diet to increased adhesiveness of formed blood elements, to biochemical alteration of central myelin, and to impaired delayed hypersensitivity are plausible mechanisms linking risk of MS to high animal-fat consumption. (Arch Neurol 31:267-272, 1974) Many investigators13 have sug¬ gested that nutrition plays a role in causing multiple sclerosis (MS). If there were a relationship between a dietary factor and the cause of MS, then one should be able to show similarities in the geographic distribution of MS and that dietary factor. In this study, we examined nutrition in various parts of the world and correlated MS frequency with the distribution of several components of diet.
MethodsInformation on nutrition has been pub¬ lished by the Statistical Yearbook of the United Nations for several decades.4 Re¬ cent dietary information from 120 coun¬ tries was used for this report and was ex¬ pressed in terms of the average daily per capita consumption of four factors: total calories, total protein (in grams), calories of animal origin (as percent of total calo¬ ries), and fats and oils (in grams). Earlier dietary data were also reviewed but were available on fewer countries. It was appre¬ ciated that even the more recent data were crude, but intensive dietary surveys in six countries (Japan, Finland, Yugoslavia, Greece, the Netherlands, and Italy) agreed well with the crude estimates for these same countries5 and enhanced confidence in the validity of the dietary data.Information on MS frequency was de¬ rived from studies of prevalence in welldefined populations." Prevalence of MS was plotted against geographic latitude and a strong positive association clearly existed (Fig 1); ie, MS tended to increase in preva¬ lence with geographic latitude. When these points were plotted by computer as a parabolic function, a good fit resulted. However, considerable scatter was evi¬ dent. Alternative mathematical distribu¬ tion models using the same data have been discussed by Alter et al.7 From these stud¬ ies, 22jttvmtrie3_WEre_£ele£ted_f ormation on both_MS._pre.valence anrLdiet was available. areas had good medical facilities and readily available information on de¬ mographic characteristics of the popu¬ lation. The case-findmg_jn_ethods in each area were fairly well standardized and in¬ volved examin^tionjrfjr^ejicjaLrficords in hospitaj ...