Monday, February 25, 2013

Excerpts from my Masters Thesis

The Effect of Fructose on Triglyceride Levels in Humans: A Systematic Review and Research Proposal

Background: The monosaccharide fructose is being investigated as a potential risk factor for cardiovascular disease, based on the premise that it causes serum triglycerides (TAG) to increase to a greater extent than glucose.
Objective: A systematic review was conducted of clinical trials comparing fructose and glucose in order to determine their respective effects on TAG levels in adults.
Design: Utilizing a literature search on MEDLINE (through May 2012), relevant controlled trials of pure fructose in comparison to glucose were examined. All such studies included a dietary fructose exposure that can be achieved through normal dietary intake.
Results: Nine of the twelve studies in this review found some evidence of a difference in the effect of fructose on TAGs, compared to the effect produced by glucose.  One of the nine trials found this result among men and not women, and one trial found that only postprandial TAGs (not fasting TAGs) were significantly increased with fructose. The remaining three studies did not find any evidence of a difference in the effect produced by fructose on TAG concentrations, compared to the effects from glucose.
Conclusions: The data in this systematic review suggest that the consumption of fructose may cause a larger increase in TAGs than the consumption of glucose. However, more research is needed on this topic due to shortcomings of studies conducted to date.

            Cardiovascular disease (CVD) is the leading cause of death in the developed world, and despite the advances in therapeutic approaches like statin drugs, rates are continuing to climb(1). Currently, 36.9% of U.S. adults have some form of CVD, which includes cardiac disease, peripheral arterial disease, vascular diseases of the kidney and brain, hypertension, heart failure, stroke, and coronary heart disease. This percentage is expected to rise to 40.5% by 2030(1).
            Atherosclerosis and hypertension have both been identified as factors that lead to the development of CVD. Meta-analyses and systematic reviews of CVD have established elevated serum triacylglyceride (TAG) levels as one of the independent risk factors for this collection of diseases(2).
            There are a variety of factors that are believed to contribute to raised TAG levels, including weight gain/obesity, a lack of physical activity, the use of tobacco, excessive amounts of alcohol, the excessive consumption of carbohydrates, diseases like type 2 diabetes and renal disorders, the use of certain drugs, and a genetic predisposition toward dyslipidemia(2).
TAGs are usually measured as part of a lipid profile, which also includes total cholesterol levels, high-density lipoproteins (HDL), and low-density lipoproteins (LDL). Occasionally an extended lipid profile may be taken which includes very-low density lipoproteins (VLDL) as well. High levels of LDL and VLDL, both of which contain large amounts of TAGs, indicate the presence of hyperlipidemia and are well established as risk factors for not only CVD but also pancreatitis and stroke(3).
Hypertriglyceridemia is a very common form of dyslipidemia in our population today(4). Generally, normal levels are considered less than 150 mg/dL, borderline high is 150-199 mg/dL, high is 200-499 mg/dL, and very high levels are considered to be greater than or equal to 500 mg/dL(5). In the 1999-2004 National Health and Nutrition Examination Survey it was found (measuring fasting TAG levels) that 33% of American adults have borderline high TAGs, 18% have high levels of TAGs, and 1.7% have very high TAG concentrations(5).  
Because of the presence of large concentrations of TAGs following a meal, fasting TAG concentrations may not be the best indicator of coronary risk. Many studies have attempted to determine whether postprandial or fasting TAG concentrations are the better predictor of atherosclerosis. Recently, several studies(6, 7), as well as several reviews(8-10) indicate that postprandial triglyceride measurements, compared to fasting values, are a better indicator of risk for coronary disease.
Carbohydrates have been found to raise TAG levels more than other dietary substances. Recent studies have attempted to determine if fructose raises TAG levels more than the consumption of other monosaccharides. Fructose is a component of a variety of commonly consumed sweeteners including sucrose (table sugar), high-fructose corn syrup (HFCS), maple syrup, and honey, among others. A potential mechanism for this physiological effect may be an increase in de novo lipogenesis (DNL), whereby the synthesis of the saturated fatty acid palmitate is activated in the liver by excess fructose consumption. The resulting hepatic metabolism may culminate in an increased flow of TAGs, high in palmitate, which are packaged in very low-density lipoproteins (VLDLs)(11). Some researchers have found that the monosaccharide fructose is unique, compared to glucose, in its tendency to cause DNL(11-13), while other researchers have failed to find this differential effect(14, 15).
Fructose consumption in the US is significant, with mean consumption estimated by The National Health and Nutrition Examination Survey (NHANES) to be 54.7 g/day, which accounts for 10.2% of total energy intake(16). Approximately 41% of the total sugars in the American diet come from fructose(16). Hence, if fructose has a more adverse effect on TAG than glucose (the other primary monosaccharide in the diet) there may be significant public health implications. The purpose of this paper is to review human studies that have evaluated whether fructose increases TAG levels to a greater extent than glucose. 

In this systematic review of RCCTs and crossover trials, data suggest that the consumption of fructose may cause a larger increase in TAG concentrations than the consumption of glucose, however, the quantity of quality studies are insufficient to draw concrete conclusions about this relationship and indicates that more research is needed on this topic.
Nine of the twelve studies in this review found some evidence of a difference in effect from fructose, compared to glucose, on TAG concentrations. In general, studies that measured postprandial TAGs; used a hypercaloric diet; and were short in duration (one day exposure) more often reported a difference in effect. Other systematic reviews on the effects of fructose on TAGs in humans have come to mixed conclusions(9, 27, 28), although only one of these reviews examined the isocaloric exchange of fructose for other dietary carbohydrates(28) and none of them compared the results of fructose exposure to equal glucose exposure on TAGs.
In this review, there were only seven studies that were rated as high quality, measured postprandial TAGs, and compared the exposure of pure fructose to an equal exposure of pure glucose(11, 20-25), six of which found some evidence of a difference in effect on TAGs for fructose, compared to glucose(11, 20-22, 24, 25).
There is a proposed biologic mechanism of action whereby fructose may increase TAG levels to a greater extent than glucose (see Figure 1). The metabolisms of the monosaccharides fructose and glucose have a number of major differences. Whereas virtually every cell in the body can metabolize glucose, fructose is largely shunted to the liver by the hepatic portal vein for metabolism. The liver responds to fructose consumption by engaging in lipogenesis, manufacturing triglycerides and packaging them in lipoproteins. Of relevance to this review is that whereas glucose metabolism is inhibited by excess energy intake, through cytosolic ATP and citrate levels(11), as well as the production of leptin and insulin, fructose consumption doesn’t affect these hormones, and the metabolism of fructose isn’t believed to be regulated by levels of intake(24). Because of these distinctions between metabolism of fructose and glucose, it is important to examine the relative effect of fructose and glucose on TAG levels at different levels of intake and in the context of both isocaloric and hypercaloric diets to evaluate whether meaningful differences in effect on TAG exist. 
The research on fructose and its relative effects on TAGs, compared to glucose, has a number of shortcomings. These include relatively small sample sizes used (the largest sample in this review had only 34 subjects); doses that are on average larger than those consumed by the general population; the limited duration of exposure in the studies conducted so far, which prevents long-term effects to be known (7 of the 12 trials in this review lasted for < 1 day); the low amount of human studies on this topic; and the fact that some of the existing studies do not compare fructose with a comparable form of glucose but use starch or another glucose source instead.
Three studies in this review had design flaws that were significant enough to question their findings regarding the effect that fructose has on TAGs compared to glucose. Two studies used starch or maltodextrose as the source of glucose rather than liquid monosaccharides in identical form for both the fructose and glucose interventions. Consequently, the validity of findings from these studies may be called into question(14, 19). The study that used starch (delivered in bread) as the source of glucose found evidence of a greater effect of fructose on TAG compared to glucose(19). The study that used maltodextrose (provided as part of a liquid diet) did not find a difference in the effect of fructose and glucose on TAG(14). Since starch, as well as maltodextrins (a lightly hydrolyzed starch product) are both made up of longer chains of glucose molecules that must be lysed during digestion, it is possible that these products may take significantly longer to digest than pure glucose. This is especially true for starch, which may be only partially digested. Adding to concern with the study that used starch as the source of glucose is the use of a food matrix (bread) for delivery of the starch, whereas the fructose was delivered in liquid from. It could be speculated that the difference of effect found between fructose (delivered in liquid form) and glucose (delivered as starch in bread) in this study is attributable to this design flaw(19). The other study in this review with a major design flaw, Hudgins, et al., compared liquid fructose to liquid glucose, but the researchers failed to compare equal doses of the two monosaccharides(13). After undergoing an OGTT (75 grams glucose in liquid form), subjects were given a single bolus dose of either fructose alone, or two different ratios of both glucose and fructose together in a randomized crossover design. The fructose dose was 0.5 g/kg body weight (BW), a second dose was 0.5 g/kg BW of both glucose and fructose (F:G), and the third dose was 1.0 g/kg BW of both glucose and fructose (2X F:G). The OGTT was a similar dose of glucose as the total amount of sugar in the F:G dose for an average subject’s body weight, so that is the dose of most interest for this review. Unfortunately, no statistical data was given for the F:G dose but the researchers stated that all three doses of fructose or fructose/glucose had significant increases in total TAG. Despite these increases in TAGs, without a direct comparison of equal doses of glucose and fructose, serious limitations exist in the ability to interpret the results of this study on the difference of effect between fructose and glucose. An order effect is also of concern in this study because the OGTT (glucose dose) was always administered first.
Another important shortcoming in the literature on this topic is the lack of consensus on whether fasting or postprandial TAG measurements are a stronger risk factor for CVD. Although recent research has shown postprandial measurements to be a better indicator of atherosclerosis, historically, fasting TAG measurements were considered to be a better indicator and therefore were used more commonly as an outcome measure. This has resulted in many older studies neglecting to take postprandial measurements, and consequently there is a lack of postprandial TAG data in three of the 12 studies included in this review. A larger percentage of studies in this review measuring postprandial TAGs, compared to studies that measured fasting TAGs, found evidence of an effect of the fructose intervention on increased TAG levels compared to the glucose intervention. Since postprandial TAG measurements have been found by many reviews(8-10) to be more reliable at predicting CVD than fasting measurements, these studies may be more indicative of the effect of fructose feeding on increases in TAG levels. Of the five studies in this review that were longer term, only two of these studies measured postprandial TAGs, and both found that fructose raised postprandial TAG levels more than glucose(11, 20), although one of them, Bantle et al., with a trial lasting 6 weeks, only found this result among men and not women(20). Stanhope et al., in 2009, with an intervention period of 10 weeks, conducted the longest study in this review. They also measured fasting and postprandial TAG levels, and found a significant increase for both men and women for postprandial TAG levels, although, interestingly, this result was not observed for fasting TAGs(11).
There are two additional shortcomings in the research conducted on this topic to date. Many studies on the effect of fructose on TAGs were of very short duration. For example, 7 of the 12 studies in this review lasted for < 1 day. Since fructose exposure is usually chronic, it is necessary that future research consists of longer trials, examining the effect that fructose has compared to glucose, over longer time periods. Also, 10 of the 12 studies here used relatively high doses, comparing fructose and glucose exposures of  > 100 g/day, or up to 30% of daily energy intake. This makes it challenging to evaluate the effect that fructose has on TAG levels, compared to glucose, at more typical intake levels, like those present in average American diets. Because of this, it is important that future research examines the effects of fructose at more realistic doses and for longer durations.
In order to accumulate an adequate amount of evidence to determine whether fructose has a greater effect on TAG levels than glucose, more high-quality research is needed. It will be necessary to conduct this research with RCCTs or randomized crossover trials lasting four weeks or more, containing large enough sample sizes in order to have the necessary statistical power, measuring postprandial TAGs, and using dosages of fructose and glucose that are more typical in the standard American diet in order to determine if fructose is the primary monosaccharide contributing to hypertriglyceridemia. Additional studies are also necessary in order to determine whether hypertriglyceridemia is the primary biomarker for atherosclerosis. If future studies provide conclusive evidence that fructose increases TAG concentrations more than glucose, both public policies and dietary recommendations may need to be adjusted. By altering these guidelines, the progression of dyslipidemia and cardiovascular disease could potentially be reduced.