This review asks the question if further research on trans fatty acids and cardiovascular health is needed. We therefore review the evidence from human studies on trans fatty acids and cardiovascular health, and provide a quantitative review of effects of trans fatty acid intake on lipoproteins. The results show that the effect of industrially produced trans fatty acids on heart health seen in observational studies is larger than predicted from changes in lipoprotein concentrations. There is debate on the effect of ruminant trans fatty acids and cardiovascular disease. Of special interest is conjugated linoleic acid (CLA), which is produced industrially for sale as supplements. Observational studies do not show higher risks of cardiovascular disease with higher intakes of ruminant trans fatty acids. However, CLA, industrial and ruminant trans fatty acids all raise plasma low-density lipoprotein and the total to high-density lipoprotein ratio. Gram for gram, all trans fatty acids have largely the same effect on blood lipoproteins. In conclusion, the detrimental effects of industrial trans fatty acids on heart health are beyond dispute. The exact size of effect will remain hard to determine. Further research is warranted on the effects of ruminant trans fatty acids and CLA on cardiovascular disease and its risk factors.
Cardiovascular disease is an important cause of death. In Western countries approximately one in three people die of cardiovascular disease. Many of these deaths occur before the age of 65 years.1 A high intake of trans fatty acids increases risk factors for cardiovascular diseases and is associated with increased risk incidence of such diseases.2, 3, 4
Trans fatty acids are unsaturated fatty acids with at least one double bond in the trans configuration, which results in a straighter shape than a double bond in a cis configuration. Trans fatty acids can be divided in two groups: artificial trans fatty acids (industrial) and natural trans fatty acids (ruminant). Although trans fatty acids may be formed from n-3 and n-6 fatty acids, they do not serve any vital functions and are not essential fatty acids. Industrial trans fatty acids are formed by partial hydrogenation of vegetable and fish oils, a process that transforms oils into semi-solid fats, which are easier to process into foods. Ruminant trans fatty acids are produced in the rumens of cows and sheep and are present in dairy and meat. Finally, conjugated linoleic acid (CLA) is a trans fat first discovered in milk, which is now produced industrially as an aid in weight loss.
In this overview we summarize the evidence on trans fatty acids and cardiovascular health. Differences and similarities between industrial and ruminant trans fatty acids are discussed; only evidence from human studies is taken into account. We will first briefly discuss the acknowledged detrimental effects of industrial trans fatty acids on cardiovascular health and then focus on trans fatty acids from ruminant sources and more specifically on CLA. Finally, we will examine whether there are still important questions in this field to be solved, or whether research on trans fatty acids and cardiovascular health can be considered completed.
Sources, structure, nomenclature and intake of trans fatty acids
Industrial trans fatty acids are formed during partial hydrogenation of vegetable or fish oils with hydrogen gas and metal catalysts. Ruminant trans fatty acids are produced in the rumens of cows and sheep. They arise through biohydrogenation and/or isomerization of cis-unsaturated fatty acids from the feed by hydrogen produced during oxidation of substrates with bacterial enzymes as catalysts.5, 6 In both cases, a range of trans fatty acid isomers are formed that differ in the position of the double bond along the fatty acid molecule. The notation for such positional isomers is exemplified by vaccenic acid, C18:1 Δ11t. It has 18 C atoms and 1 double bond in the (trans) configuration at C-atom position number 11.
Industrial and ruminant trans fatty acids consist of the same positional trans isomers, but in different proportions. The isomer profile depends on conditions of hydrogenation, such as catalysts used and temperature of hydrogenation for industrial trans fatty acids and rumen pH, and the composition of oils in the diet for ruminant trans fatty acids. C18:1Δ10t and Δ9t (elaidic acid) are typically the main isomers of industrial trans fat, and vaccenic acid (C18:1 Δ11t) is usually the main component of ruminant trans fat. However, industrial trans fats also contain a considerable amount of vaccenic acid.7
CLA is another trans fatty acid. CLA has two double bonds, one in the trans and one in the cis configuration. The double bonds in CLA are conjugated, which means that CLA has only one carbon atom between the two double bonds instead of the usual two. The two main isomers of CLA are cis-9, trans-11 CLA and trans-10, cis-12 CLA. Cis-9, trans-11 CLA is naturally present in small amounts in ruminant fat. Supplements are a much larger dietary source of CLA. CLA in these supplements is manufactured from vegetable oils8, 9 and contains both cis-9, trans-11 CLA and trans-10, cis-12 CLA.
In this review, we follow the chemical nomenclature and designate all fatty acids with at least one double bond in the trans configuration as trans fatty acids or trans fats for short. However, the Codex Alimentarius standard and some countries do not define animal trans fats and trans fats with conjugated double bonds as trans fats in a legal sense, because they are thought to lack the harmful effects of ‘real’ trans fatty acids. Thus, in those countries CLA does not have to be labeled as trans fat on a food product. The government of Denmark has put a limit on trans fat in foods, but exempts ruminant trans fats.
The intake of trans fatty acids has decreased considerably over the past two decades because the food industry has largely eliminated industrial trans fatty acids from foods. This has resulted in average intakes in European countries of 1–2 energy percent10 and less than 1 energy percent in the United Kingdom, with a major proportion coming from dairy and meat.11 Intakes in the USA are coming down as well, but not as much as in Europe.12 In Europe and the USA the average intake of trans fatty acids from ruminant sources is around 0.5 energy percent.10, 12
Industrial trans fatty acids and cardiovascular health
In the early 1990s metabolic studies showed that consumption of industrial trans fatty acids raised low-density lipoprotein (LDL) and lowered high-density lipoprotein (HDL) cholesterol levels in the blood.13, 14 Furthermore, observational studies showed that a higher intake of trans fatty acids was associated with a higher risk of coronary heart disease.15 In 2006, results of observational studies were pooled and that resulted in a multivariable adjusted relative risk of 1.23 (95% confidence interval 1.11–1.37) for coronary heart disease for isocaloric replacement of 2 energy percent of carbohydrate by trans fatty acids.16
Thus, both observational studies on the incidence of cardiovascular disease and metabolic intervention studies on lipoproteins indicate that industrial trans fatty acids have detrimental effects on cardiovascular health.4, 15, 16, 17 Based on these results, industries and governments have taken actions to limit the intake of trans fatty acids from industrial sources.18, 19 There is still debate about the size of the effect of industrial trans fatty acids on cardiovascular disease and about effects of industrial trans fatty acids on metabolic pathways other than lipoproteins that may lead to cardiovascular disease. Still, the detrimental effects on cardiovascular disease are scientifically established, and it is unlikely that industrial trans fatty acids will be brought back into the food chain. At present, the debate has shifted towards the effects of ruminant trans fatty acids and CLA.
RUMINANT Trans fatty acids, and CLA and cardiovascular health
Effects on cardiovascular disease
Investigating effects of ruminant trans fatty acids on disease outcomes and comparing these effects with industrial trans fatty acids is challenging because ruminant trans fatty acids are only present in small amounts in the diet. Until the mid 1990s partially hydrogenated vegetable oils containing up to 50% trans fatty acids were widely used in the food industry. In contrast, milk fat naturally contains only some 5% trans fatty acids. As a result, observational studies have lacked the power to pick up an effect of this small amount of ruminant trans fatty acids, if any, amid so many other factors that affect the risk of cardiovascular disease. Many observational studies have investigated the associations between trans fatty acid concentrations in blood or adipose tissue and the risk of cardiovascular disease, but when blood or adipose tissues are studied it is not possible to distinguish which trans fatty acid is coming from which source. Only very few observational studies have investigated separate effects of dietary ruminant trans fatty acids. A meta-analysis of four prospective cohort studies did not show a significant association between intake of ruminant trans fatty acids and coronary heart disease (relative risk=0.92; 95% confidence interval 0.76–1.11). Intake of total trans fatty acids in these studies ranged from 2.8–10 g/day.20
One could argue that the question whether ruminant trans fatty acids cause cardiovascular disease is irrelevant, because their intake is so low.21 This holds even more so for CLA from natural sources, because CLA constitutes a tiny fraction of ruminant trans fatty acids. However, intake of CLA in the form of supplements can be appreciable, recommended dosages are up to 6 g/day.
Effects on lipoproteins: updated review
In 2010, we reviewed available metabolic studies on effects of ruminant and industrial trans fatty acids on HDL and LDL cholesterol level in humans.4 We now updated our review and added data on 4 more studies on ruminant trans fatty acids and 15 more studies on CLA. Other differences with our previous review are that we now weighed the studies for study size, we compared the effect of the treatment with the original control treatment of the study, instead of recalculating all studies to cis monounsaturated fatty acids, and we report the results in grams of fatty acids instead of energy percentages.
We combined data from different studies using linear regression, with intake of trans fatty acids as the independent variables and change in LDL cholesterol, HDL cholesterol, the plasma LDL/HDL cholesterol ratio and the total/HDL cholesterol ratio as dependent variables. We calculated these ratios on the basis of mean values of treatment differences within studies, therefore, we lacked an estimate of variation for these ratios. To maintain uniformity, we recalculated the ratios of LDL/HDL cholesterol and the ratios of total to HDL cholesterol from the mean LDL, HDL and total cholesterol levels for all studies, even when ratios had been reported. We weighed the studies for size using the inverse of the square root of the number of participants. We used comparisons with the original control treatments in all cases.
The regression lines calculated with the conventional least square methods showed a positive intercept with the y-axis. This may be an artefact because conventional linear regression algorithms presuppose that x-values have zero error.22 However, not every meal or capsule consumed in a CLA trial is precisely analyzed for CLA, and therefore reported CLA intakes contain random errors. As a result the slope calculated with conventional regression software is biased downwards, and the y-intercept goes up. Correction of this error requires knowledge of the size of the error in the intake variable, but that is unknown. Forcing the regression line through the origin is a suitable alternative because a zero change in diet should produce a zero change in blood lipids. Treatment with a zero dose of CLA as such does not necessarily produce a zero change, because enrolling subjects in a trial may affect their lipid levels. However, all data points analyzed represent the effect of a treatment with a certain amount of CLA minus the effect of a concurrent control treatment that contained a different amount of CLA. If the test and control treatment both contain 0 g of CLA, then logic allows no other outcome for the change in a lipid value than 0 plus or minus a random error term, and a regression line that does not contain the (0,0) point contradicts logic.
Ruminant trans fatty acids
We identified a total of eight published and one unpublished study on effects of ruminant trans fatty acids on lipoproteins; these include 11 comparisons and 672 subjects (Table 1). The unpublished results are from a study23 that has not yet been published in a peer reviewed journal (September 2012). Therefore, we used the values as presented at the International Dairy Conference in November 2010 (http://www.wds2010.com/delegates/presentations/10wed/06-Session0_2-David%20Baer.pdf). Linear regression showed that replacement of 1 g of control fat by ruminant trans fat increased the plasma LDL/HDL ratio and increased LDL cholesterol (Table 2; Figure 1). Results were similar to our earlier review except for the effect on HDL, for which we do not show a significant effect now. This apparent discrepancy is explained by addition of new studies and by the use of results of the original control treatment instead of recalculating effects to cis monounsaturated fatty acids. Specifically, one study that used industrial trans as control treatment and provided a raise of HDL on ruminant trans fatty acids of 0.049 mmol/l/g in the current review,24 whereas comparison with cis monounsaturated fatty acids would lead to a decrease in HDL of −0.06 mmol/l HDL/g.
We found 32 papers that investigated the effect of CLA on lipoproteins (Table 1); these included 47 comparisons and 2048 subjects. Replacing 1 g of control fat by CLA in the linear regression model increased LDL cholesterol, decreased HDL cholesterol, increased plasma LDL/HDL ratio and increased the total cholesterol/HDL ratio (Table 2). An intake of 3 g of CLA, as advised as minimal amount by most producers, would be expected to lead to an increase in LDL cholesterol of 0.045 mmol/l (Table 2).
Weight loss affects lipoprotein concentrations. Therefore, we also calculated the effect of CLA if we excluded studies that were intended for weight loss or where subjects had a change in body weight of more than 1.5 kg. This left 23 studies with 30 comparisons (Figure 2; Table 1). The linear regression model for those studies again showed that CLA significantly increases LDL cholesterol, the plasma LDL/HDL ratio, and the total cholesterol/HDL ratio (Table 2).
Gram for gram, ruminant trans fatty acids, CLA and industrial trans fatty acids had largely the same effect on blood lipoproteins. We therefore suggest that fatty acids with a double bond in the trans configuration unfavorably affect blood lipids and lipoproteins, regardless of whether they have been produced in factories or in the rumens of cattle and sheep.
Other pathways through which trans fatty acids may influence cardiovascular health
One of the unsolved questions in this research field is what the exact effect is of trans fatty acids on cardiovascular disease. The risk estimates from the observational studies are much larger than can be explained by the effect on lipoproteins only. This could be due to residual confounding, but there could also be additional adverse effects of trans fatty acids on cardiovascular disease through pathways other than lipoproteins.
Other suggested pathways are systemic inflammation, oxidative stress, endothelial function and insulin resistance or diabetes. Several studies, but not all, suggest that industrial trans fatty acids have unfavorable effects on inflammation.25 For inflammatory effects of ruminant trans fatty acids and CLA there are not many indications that they have strong effects on inflammatory processes.26, 27, 28, 29, 30 However, the available number of studies is small and the dosages of ruminant trans fatty acids used in these studies are low.
Effects on other pathways are much less clear. For example, future research has to show if trans fatty acids affect the glucose–insulin homeostasis or diabetes. Unfavorable effects on this pathway would also have consequences for the magnitude of the effect of trans fatty acids on cardiovascular disease. It could partly explain why the association between trans fatty acids and cardiovascular disease is stronger than predicted from changes in lipoproteins. However, until now, results of trials and observational studies on diabetes risk or on indicators of diabetes are equivocal.25, 30
Summarizing resolved and unresolved questions: all trans fatty acids have detrimental effects on lipoproteins. That includes HDL when trans fatty acids are compared with cis fatty acids.4 Effects on other pathways are less clear. Avoidance of trans fatty acids is likely to reduce the risk of cardiovascular disease. Changes in the formulation of foods, labeling of food products and actions from governments have resulted in a dramatic decrease in the intake of trans fatty acids from industrial sources.18, 19 The intake can be further decreased if industrial trans fats are avoided in the preparation of fried foods and bakery shortenings.31
The average intake of trans fatty acids from ruminant sources is only about 0.5 energy percent. The effect of animal trans fatty acids on cardiovascular disease risk is still unclear, but the impact if any must be minor and removal of these trans fatty acids from milk and meat is technically not feasible. However, consumers who follow the long-standing advice to choose low fat dairy products will automatically also reduce their intake of ruminant trans fatty acids.
Intake of CLA can be substantial because CLA is sold as supplements. We conclude that CLA has unfavorable effects on lipoprotein levels similar to those of other trans fatty acids. It is unclear what the effects of CLA are on markers of diabetes. The Food Standards Australia and New Zealand bi-national government agency decided on this basis to ban foods enriched with CLA from the market (http://www.austlii.edu.au/cgi-bin/sinodisp/au/cases/cth/AATA/2012/551.html?stem=0&synonyms=0&query=title(Food%20Standards%20)). The Decision of the Tribunal that upheld this policy gives a good overview of the potential adverse effects of taking CLA supplements on a regular basis.
Industrial trans fatty acids have been shown to have detrimental health effects. Therefore, future research should continue to find alternatives for these fats that have similar baking properties, but not the adverse health effects. This would specifically be helpful for the bakery industry. Thus, further research on industrial trans fatty acids and cardiovascular health is not a top priority.
The proportion of ruminant trans fatty acids in foods is low and will further decrease if consumers follow the advice to decrease intake of saturated fatty acids. Therefore, ruminant trans fatty acids are not an urgent research topic either. Even though there are gaps in our knowledge of their effects. Future research is warranted to provide more precise insights into effects of CLA on both lipoprotein concentrations and on inflammatory processes, and insulin and diabetes related factors.
Our studies on trans fatty acids and lipoproteins were supported by the Netherlands Heart Foundation (Grant No. 2006B176) and the Foundation for Nutrition and Health Research. MBK was supported by an Academy Professorship of the Royal Netherlands Academy of Arts and Sciences. AJW was partially supported by Food Standards Australia New Zealand for updating statistical analyses. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
About this article
Journal of the American Oil Chemists' Society (2015)