Fish consumption, blood docosahexaenoic acid and chronic diseases in Chinese rural populations

Abstract

The Chinese traditional diet is low in fat. However, there is regional variability in the amount, type of fat consumed and the pattern of chronic diseases. An epidemiological survey of 65 rural counties in China (6500 subjects) was conducted in the 1980s. We have re-examined the red blood cell fatty acid and antioxidant composition, with fish consumption. Fish consumption correlated significantly with the levels of docosahexaenoic acid (DHA) in red blood cells (RBC) (r=0.640, P<0.001), selenium (r=0.467, P<0.001) and glutathione peroxidase (r=0.333, P<0.01) in plasma. The proportion of DHA in RBC was inversely associated with total plasma triglyceride concentrations. A strong inverse correlation between DHA in RBC and cardiovascular disease (CVD) was found. The strongest correlation was the combination of DHA and oleic acid. RBC docosahexaenoic acid itself also correlated negatively and significantly with most chronic diseases and appeared to be more protective than either eicosapentaenoic or the ω3 docosapenataenoic acids. These results demonstrate the protective nature of fish consumption and DHA, found in high fat Western diets, operates at a low level of fat. This finding suggests the protective effect of fish consumption as validated by red cell DHA is universal. The protective effect is, therefore, most likely to be due to the fundamental properties of docosahexaenoic acid in cell function.

Introduction

The massive decrease in Chinese death rates from nutritional deficiencies, infectious and parasitic diseases has been one of the most important, practical medical successes in the last few decades. We collaborated with Chen Junshi, Colin Campbell, Richard Peto and others in a study of the health characteristics of 65 Chinese counties (6500 subjects) entitled Diet, Life-Style and Mortality in China (Chen et al., 1990). The outcome publication contained 894 pages of data, maps and text describing the geographic differences in disease. The authors were anxious that the database should be made use of by anyone with an interest in epidemiology. Epidemiological correlations on their own are meaningless unless there is an evidence base. Epidemiological data no matter how robust, is only a set of correlations, some of which may be chance, a surrogate for some other factor or meaningful in the context of causation. To separate chance and meaningful data, the identification and the test of a potential mechanism are required, all of which takes much time. The significance of the 65-County Study in China is that the study was not simply a correlation of one factor, in this case, diet with disease, it also included evidence on objective, nutritional parameters enabling the primary outcome to be tested further in terms of the secondary variables and chronic disease. Although attracting less attention than the epidemiology of diseases affecting Western populations, the study in China is of equal importance in terms of the number of people to whom public health measures may make a difference.

The secondary variables included trace elements, vitamins and essential fatty acids (EFA). With a wealth of data now available from animal studies on the effects of different fatty acids on cell function, these secondary variables become of considerable significance both to food policy in China and also has implications for the rest of the world. From the database came several conclusions as to relationships between nutrients and the principal chronic diseases in children and adults. This design short-circuited the search for plausible mechanisms for differences in disease patterns in different populations.

The authors, however, warned that the results need to be assessed with caution, as would be the case of any large epidemiological study. Apolipoprotein B included in the study, is a marker of the type of blood cholesterol that is associated with coronary heart disease (CHD) in the countries such as the US where the intake of dietary saturated fats and trans isomers from land animal products and their analogues is high. In China, the fat intake comprised approximately 15% of the energy compared to the US with 40–45%. The Chinese mean plasma cholesterol (mean=127 mg/dl) is little more than half of that seen in the UK or US. Examination of the variation of the low incidence of CHD in different Chinese counties revealed that apolipoprotein B was only weakly correlated with mortality from CHD. This may mean that different factors operate in the context of a low fat diet compared to a high fat diet. If for example, the ω3 fatty acids were protective in the context of the low fat Chinese diet, it would suggest that these fatty acids have an intrinsic value in their own right and are not just counteracting the effects of a high fat diet. That intrinsic value would most likely be concerned with the integrity of the vascular cell membranes and regulation of blood flow and signalling.

Currently, the discussion on dietary aetiology of CVD mainly concerns cholesterol, saturated fats and trans isomers. These are all negative or destructive forces which have focused interest on harmful effects of too much fat. However, the Greek diet, which also seems to be protective, is just as rich in fats as the usual Western diet of the UK and USA. The difference is that the Greek fat is mainly olive oil, which has a low saturated fat and high monounsaturatred fatty acid content (Assmann et al., 1997). This contrast in diet associated with their low mortality from CVD led to the incrimination of the nature of the fat not just the amount and specifically to the deleterious effects of the high saturated fat diet (Keys, 1997). In this sense, the Chinese study represents a squaring of the Seven Country Study by adding a large population on a low fat diet whereby the nature of the positive fats would be revealed. A low fat diet might provide too little of the essential fats needed for the health of the vascular system. Our two previous analyses of this database only discussed the associations between the fatty acids in RBC fatty acids, antioxidant status and mortality of CVD (Fan et al., 1990, Hensrud et al., 1994). However, there is extensive data on chronic diseases in the book by Chen et al. (1990). In this paper, we report the red cell fatty acid relationships with several of these other health statistics.

The lipid bi-layer of the cell membrane is subject to nutritional and environmental stress. Its highly unsaturated component fatty acids are very susceptible to peroxidation. We have evidence that a high saturated fat diet is likely to alter the membrane lipids and vascular function (Ghosh et al., 2001). There is thus good reason to raise the hypothesis that membrane related disorders, such as stroke, inflammation, immune disorders, arteriosclerosis and its sequelae in CVD, are a consequence of distorted membrane lipids brought about by inadequate dietary lipids, peroxidative stress and genetic variations. Indeed, the late Hugh Sinclair suggested arteriosclerosis was an essential fatty acid deficiency (1956) in which he included a DHA deficiency (Sinclair, 1956, Sinclair, 1968).

Epidemiological studies on heart disease, consistently indicate a protective role for fish and sea food consumption as opposed to land animal fats. The traditional Eskimo, Japanese and Mediterranean diets are rich in sea foods and there is also evidence of fresh water lake foods providing protection (Marcovina et al., 1999). Indeed, fish consumption is claimed to be associated with a reduced risk from all-cause, ischemic heart disease and stroke mortality at the population level (Bang and Dyerberg, 1987, Bulliyya et al., 1997, Zhang et al., 1999, Gillum et al., 2000, Oomen et al., 2000, Torres et al., 2000). Clinical trials have provided further support for a protective role of fish oil fatty acids. The DART study (Diet And Reinfarction Trial) (Burr et al., 1989) tested fibre, a low fat diet and fish or fish-oils. The results were negative for the first two but in the fish or fish oil group, mortality fell rapidly with 29% less deaths after 2 years. The GISSI trial also produced impressive results in prevention (GISSI—Prevenzione Investigators, 1999) with much data supporting a health protective role for fish oils (Simopoulos, 1991, FAO and WHO, 1994, Nordoy, 1999, Simopoulos, 1999, Connor, 2000, Franzosi et al., 2001). Moreover, a mechanism for the anti-arrhythmic effects of DHA has now been described which would explain the rapid and striking reduction in sudden death found in the clinical trials (Leaf et al., 1999). However, all of the above evidence has been obtained against a background of a high fat diet. Hence, in the light of the accumulation of knowledge of the role of DHA in health, we have re-examined our data on red cell polyunsaturated fatty acids in the Chinese study and wish to report on the geographical relationships with the wide and contrasting disease pattern in the study database. DHA is the most polyunsaturated of the usual dietary and membrane fatty acids. In view of the concern some people express about the highly unsaturated fatty acids imposing a peroxidation threat we also report the impact of red cell DHA and fish consumption on antioxidant status. In addition, we have re-examined the association between DHA in RBC and approximately three-dozen mortality rates of non-cancer disease in this database. We report here a negative correlation with DHA status to most chronic diseases. In contradiction of a commonly held view of an association between highly unsaturated fatty acids and peroxidation, we also wish to report a positive correlation between DHA and anti-oxidant status.