Search

Linoleate, linolenate, arachidonate, docosahexaenoate and six other fatty acids were major components of 24 ester preparations fed as 5% of the diet for 60 days to groups of male white rats. The experiment was designed so as to provide that all major fatty acid components were independent of each other in the sense that the intake of each was poorly correlated with the intake of any of the others. Fatty acid compositions of liver lipids were determined and were related to the composition of the diet lipids. Linolenate and docosahexaenoate contents of diet and tissue revealed the same relationships reported previously from experiments in which individual pure acid esters were added to a fat-free diet. Linoleate, when fed in lipid mixtures, was more effective in raising the linoleate concentration in liver lipids than when fed alone, but this increase did not change the shape of the dose-response curve or the estimated nutritional requirement. Large amounts of fish oil in the diet tended to depress the arachidonate concentration in tissue lipids.

Eight saturated fatty acid esters were fed to male white rats for 30 days in a 1/81 fractional factorial experiment in which diets contained 12–38% of their total energy as lipid. Marked increases in food intake, feed efficiency, and weight gain were achieved when lipid provided 36% of diet energy, and when that lipid was more than half caproate, caprate, myristate, and/or stearate. Caproate was the only saturated fatty acid to increase the concentration of cholesterol in plasma and liver. Caproate also increased plasma glucose levels. The feeding of stearate or caprylate decreased plasma and liver cholesterol. Caprate increased liver fat. The short-chain fatty acids (butyrate to myristate) increased the concentration of fat in the carcass.

Twelve young men were fed controlled diets containing three known levels of linoleate in the fat for periods of fifteen days. At the end of each diet period blood was drawn, and the serum lipids were extracted. The fatty acids of total lipids, triglycerides, phospholipids and cholesteryl esters were analyzed for their dienoic, trienoic and tetraenoic fatty acids. The algebraic sum (dienoic − trienoic + tetraenoic) was plotted against log linoleate intake, and the equations for the linear relationships were derived by the method of least squares. The best estimate of dietary linoleate was found to be the (dienoic − trienoic + tetraenoic) acids of serum triglyceride fatty acids. However, the comparable relationship for total fatty acids and dietary linoleate gave such high correlation between these variables (r = 0.90) that fractionation of serum lipids seems unjustified. The equation for this relationship is log10linoleate = −0.296 + 0.0243 (diene − triene + tetraene). This relationship is offered as a means of estimating dietary linoleate from serum analysis. The correlation between dietary linoleate and total serum cholesterol was found to be −0.24.

The content of polyunsaturated fatty acids in serum total fatty acids was related to the dietary intake of linoleate in 243 infants two to four months of age and in 197 infants eleven to twelve months of age. The curves relating triene:tetraene ratio to dietary linoleate indicated approximately 1 per cent of total calories as minimal linoleate requirement. The exponential equations relating dienoic, trienoic and tetraenoic acids to dietary linoleate were derived by computer methods. These yielded constants from which the minimal nutrient requirement could be calculated. The best estimates were obtained from the triene and tetraene data for two to four month old infants and were found to be approximately 1.4 per cent of the caloric intake.

A number of statistical techniques were applied to the study of tissue analysis, dermal score, and weight gain data obtained from rats maintained with controlled intakes of linoleic acid and total fat. The correlation between each of these variables and dietary linoleate as well as all other variables was presented. The correlation matrix thus generated was used for multiple regression and factor analysis studies.

To test whether effects and interactions of fatty acids when fed singly also apply in complex dietary mixtures, a study was undertaken using natural fats and mixtures. Twenty-one dietary fatty acid ester mixtures were prepared from common fats and oils plus small amounts of purified fatty acid esters. The compositions of these mixtures were adjusted to allow independent but simultaneous study of twelve of the dietary fatty acids. The lipid mixtures were fed as 10% of the dietary calories to groups of weanling male rats for 66 days. The rats were killed and the liver lipids extracted and analyzed for fatty acid content by gas chromatography. Metabolic interactions and conversions were indicated by intercorrelations calculated with the aid of a digital computer. Equations for estimating the amounts of 4 dietary fatty acids from analysis of tissue, and of 7 tissue fatty acids from an analysis of diet are presented. The linolenate requirement of the rat is estimated from these latter equations to be 0.14% of caloric intake. There are high, positive diet-tissue correlations between acids of the linoleate family and between acids of the linolenate family. The amounts of certain saturated and monoenoic fatty acids in the diet had marked effects upon these correlations.

A new method of estimating minimum nutrient requirements is proposed which is particularly well suited to the interpretation of tissue analysis or weight gain data. For the first step of this procedure, exponential equations were fitted to the curves relating tissue PUFA concentration to dietary linoleate concentration. Constants derived from these equations were then used to demonstrate that there are 4 distinct metabolic processes related to linoleate nutrition. Two of these were found to correspond to linoleate requirements of 0.6 and 1.5% of calories, respectively. The former of these (0.6 Cal./100 Cal. of diet) was associated with the decrease in tissue diene and increase in tissue triene that is observed as linoleate intake decreases. The latter (1.5 Cal./100 Cal. of diet) was associated with tetraene metabolism in the tissues, and total body weight gain. The other two metabolic processes were not interpreted in terms of nutritional requirements. One was related to the storage of excesses. The other was associated with tissue lipid changes observed at linoleate intakes below 0.4% of calories, and might be referable to variations of linolenate in the diets used in this experiment. These latter changes appear complex and require further study before interpretation.

Available statistical methods that, by digital computer technics, can now be applied quickly and easily to large volumes of data, allow one to take the usual food consumption data for a group of individuals and to derive from it a fairly clear picture of the eating habit patterns of the different individuals within that group. They allow one to locate the specific individuals within the group whose eating habit patterns differ in a significant way, and to point to the nature and extent of these differences. They also provide suggestions concerning the planning and interpretation of dietary survey data.

Assumptions implicit in the use of pair-feeding techniques have been discussed, and the logical consequences of using the pair-starved control as a basis for interpreting the potassium loss of radiation sickness have been explored. This type of experimental design assumes that the metabolic effects of starvation are major factors in determining the over-all experimental result. In the example studied this was not true. When this a priori assumption is not fully justified, erroneous conclusions result, and there is a loss of the total data. In many cases, therefore, the non-critical use of pair-feeding techniques may constitute a distinct hazard.

The micromethod of Besseyet al.1for serum vitamin A has been shown to be subject to a number of errors. The primary difficulty arises from the fact that ultraviolet light not only destroyed vitamin A but changed other substances that absorb at 328 m. The results secured by the above method did not agree with those from another spectrophotometric method. The differences were especially large for those serum samples containing low levels of vitamin A.