Is Cottonseed Oil the Next Healthy Oil?

Abstract

Cotton is primarily recognized as a textile crop; however, recent evidence suggests that cottonseed oil (CSO) may be a simple and effective dietary approach to improving cardiometabolic risk factors. The purpose of this perspective is to draw attention to the current literature examining the physiological benefits of CSO consumption and highlight the remaining questions that need to be answered to fully evaluate the potential of CSO to be the next important nutrition intervention for improving cardiometabolic health.

INTRODUCTION

Cottonseed oil (CSO) has been in the food supply for over 100 years. Despite public criticism for its prominent role as a textiles plant, cotton is regulated by the Food and Drug Administration as a food crop due to its byproducts, namely CSO. Commonly utilized in ready-to-eat foods such as salad dressings, vegetable oil blends, and prepackaged snacks (cookies, crackers, potato chips), CSO is present in today’s food supply. As a cooking oil, CSO offers a high smoke point, making it suitable for use in high heat cooking, and its neutral flavor does not mask other food flavors. Further, CSO has a high antioxidant content, primarily alpha- and gamma- tocopherols, ¹ which improve the stability of products made with this oil. CSO has a unique fatty acid profile, including a high level of (51.9%) polyunsaturated fatty acids (PUFA), most of which is the omega-6 (n-6) PUFA, linoleic acid (LA) (51.5%). However, CSO also contains moderately higher amounts of saturated fatty acids (SFA; 25.9%). ² With this fatty acid profile, it may be surprising that CSO is gaining interest as a “healthy oil”.

While CSO may not be the first oil that comes to mind when discussing “healthy oils”, its ability to lower cholesterol has been observed in preclinical models dating back to the 1950s. The first observations by Aftergood et al., ³ and Okey et al. ⁴ showed dietary CSO lowered cholesterol in rats compared with lard and coconut oil, both high sources of SFA. Later in the 1980’s, 90’s and early 2000’s, Hampden et al., ⁵ Edwards et al., ⁶ and Radcliffe et al. ⁷ showed CSO to lower cholesterol compared with corn oil in male rats, later confirming the effect in females. ⁸ Through this work, CSO was shown to be sufficient to lower cholesterol compared with another PUFA-rich oil (corn oil). This interesting result suggested CSO not only had lipid lowering effects compared with SFA-rich sources but maybe had superior lipid lowering effects compared to other PUFA-rich oils. This led to speculation about something unique in CSO beyond just its fatty acid profile (high PUFA) exhibiting superior lipid lowering effects. These interesting preclinical observations in animal models also spurred the investigation of CSO enrichment of the human diet.

COTTONSEED OIL IMPROVEMENTS IN BLOOD LIPID PROFILES IN HUMANS

To date, CSO diet enrichment has been assessed in healthy young adults ^9,10 and older adults with hypercholesterolemia. ¹¹ Clinical trials providing high quantities of CSO (average of 75–95 g CSO per day) have been shown to consistently lower fasting total cholesterol (TC) and low-density lipoprotein cholesterol (LDL-c). ^{9 –11} The magnitude of these lipid lowering effects range from 5.2% to 15.3% reductions in TC and LDL-c in healthy populations. ^9,10 In adults with hypercholesterolemia, Prater et al. ¹¹ found TC and LDL-c were reduced by 7.4% and 12.2%, respectively. These effects were strong enough that 57% of participants in the CSO group experienced a reduction in at least one diagnostic category for TC and/or LDL-c. ¹¹ Additional measures of lipoprotein profiles from this study, including non-high-density lipoprotein cholesterol (non-HDL-c) and apolipoprotein B, were also significantly reduced by an average of 11.9% and 10.5%, respectively. ¹¹

The effect of these CSO interventions on fasting HDL-c and triglycerides (TG), however, has been shown to be transient. ^{9 –11} Polley et al. ¹⁰ observed a 29.6% reduction in TG and a 7.6% increase in HDL, while Prater et al. ¹¹ observed no change in TG but a 6.7% increase in HDL that was not different from the comparison group. Furthermore, Davis et al. ⁹ observed no changes in TG or HDL-c with CSO diet enrichment. Despite these inconsistent effects in HDL-c and TG, the consistent TC and LDL-c lowering effect observed with CSO consumption is of large enough magnitude to potentially elicit clinically meaningful cardiovascular disease risk reduction. ¹¹

Two of the CSO feeding trials also studied lipid responses in the postprandial state. In young healthy males, Polley et al. ¹⁰ provided meals rich in CSO or olive oil (OO; the comparator oil in this trial) and found significant improvements in postprandial TG responses after the high CSO vs. OO diet. ¹⁰ These results showed that daily consumption of CSO in healthy males was sufficient to elicit improvements in lipemia in as little as 5 days. Conversely, Prater et al. ¹¹ used a high SFA meal challenge, which did not result in a lowering of postprandial TG after 8 weeks of CSO v. OO diet enrichment, but rather a worsening TG response after the OO diet only. Further, there was a trend for suppressed non-esterified fatty acids with CSO compared with OO enriched diets. ¹¹ The participants were in a slight energy surplus, so results indicated that CSO may offer protection to postprandial lipemia even under the metabolic stress of energy surplus in adults with hypercholesterolemia.

POTENTIAL MECHANISMS FOR COTTONSEED OIL IMPROVEMENTS IN LIPID PROFILES

Potential mechanisms explaining the improvements in blood lipid metabolism have been explored in pre-clinical models. Paton et al. ¹² demonstrated that CSO enriched high-fat (HF) diets did not result in the same liver lipid accumulation as observed for high SFA or safflower oil diets (high PUFA diet). This work showed that CSO contains small quantities of a bioactive lipid, dihydrosterulic acid, that in conjunction with the LA content of CSO, suppressed the activity of stearoyl-CoA desaturatse-1 (SCD1). SCD-1 is a Δ9-desaturase responsible for the desaturation of SFA to produce monounsaturated fatty acids for storage in lipid droplets. ¹³ The inhibition of SCD-1 has been an anti-obesity target sought unsuccessfully through pharmacologic methods, ¹² which highlights the novelty of the mechanism through which CSO acts. Accumulation of lipid in the liver, such as in cases where SCD-1 is active and provided significant substrate (HF diets), is known to impair liver health and result in hyperlipidemia. Polley et al. ¹⁰ confirmed the possibility of CSO inhibition of SCD-1 in healthy males through measurement of a desaturation index in plasma. This same technique was used in Paton et al. ¹² to suggest SCD-1 inhibition, but further mechanistic measures in liver tissues were able to confirm this mechanism in the mouse model. The improvement in liver health with CSO has been further confirmed in broiler chickens, and additionally shown to improve total antioxidant capacity. ¹⁴ While this work has shown the potential of CSO to protect the liver from lipid accumulation in healthy models, subsequent work was not sufficient to completely reverse non-alcoholic fatty liver (NAFLD; pathological lipid accumulation in the liver) in a mouse model. ¹⁵ Despite the lack of complete reversal of NAFLD, the CSO diets improved plasma and liver lipid profiles, potentially through increases in peroxisomal lipid oxidation via ACOX1. ¹⁵ While this work shows that CSO is acting at the liver to improve lipid metabolism, more work is needed to fully elucidate the mechanisms through which it is achieving liver improvements and subsequent improvements in lipoprotein metabolism.

THE QUESTION OF INFLAMMATION

Nearly all of the PUFA content in CSO is the n-6 PUFA, LA. Due to hesitations around hypothesized pro-inflammatory properties of dietary n-6 PUFAs (specifically LA), inflammatory marker responses following CSO consumption have been investigated. Two studies have analyzed the inflammatory cytokine responses of HF, CSO-rich diets compared with OO-rich diets and did not find any evidence of increases in inflammatory markers (C reactive protein [CRP], interleukin-6 [IL-6], interleukin 1-β [IL-1β], and tumor necrosis factor-α [TNF-α]) with either diet. ^16,17 Further, in the trial of young healthy males, TNF-α was actually reduced with the CSO diet. ¹⁶ Taken together, these data suggest that a CSO enriched diet does not appear to elicit an increase in select markers of inflammation. While one of these trials was performed in adults with overweight/obesity who are at risk for low-grade inflammation, ¹⁷ it should be noted that the populations tested did not have other conditions associated with severe inflammation such as cancer or rheumatoid arthritis. It is yet unknown how these populations would respond to CSO diet enrichment.

INFLUENCE OF COTTONSEED OIL ON OTHER CARDIOMETABOLIC RISK FACTORS

While CSO consumption has been primarily studied in the context of lipid metabolism, its effects on other factors that influence energy balance, namely metabolism and appetite have also been assessed. In young adult males, 5 days of CSO diet enrichment increased postprandial fat oxidation after CSO-rich meals; ¹⁸ however, no improvements in postprandial fat oxidation were observed in older adults with hypercholesterolemia after 8 weeks of CSO consumption when given high SFA test meals. ¹⁹ Further, there were no changes in fasting or postprandial energy expenditure with CSO in either investigation. ^18,19 On the other side of the energy balance equation is energy intake which is driven by appetite. Only two studies have assessed appetite responses to diets enriched with CSO, and both used OO as the reference group. ^20,21 Acute CSO-rich meals were shown to stimulate a greater postprandial cholecystokinin response (CCK; fullness signaling hormone) and a greater suppression of postprandial ghrelin (hunger signaling hormone) compared with OO-rich meals. Despite both physiologic and perceived differences in appetite measures after acute meal challenges, there were no differences in the direct measure energy intake at a buffet meal. ²¹ Further, 5 days of CSO-rich diets resulted in suppression of hunger ratings compared to OO diets, but again, this did not lead to greater reductions in energy intake at a buffet meal. ²¹ Similarly, adults with hypercholesterolemia consuming 8 weeks of CSO-rich diets displayed greater increases in fasting CCK, greater suppression of post SFA meal ghrelin, and greater increases in postprandial feelings of fullness. ²⁰ These findings are robust on their own, but were further punctuated by greater reductions in self-reported ad libitum energy intake compared with similar OO-rich diets. ²⁰ Taken together, daily CSO consumption may have an appetite suppressing effect, which could be important in energy balance and weight maintenance.

Lastly, CSO has been shown to improve other biomarkers that may influence cardiometabolic health. These include angiopoietin-like proteins (ANGPTLs), which are metabolically active regulatory proteins, blood glucose levels, and/or markers of coagulation potential. For ANGPTLs, three specific isoforms have been studied: ANGPTL-3, -4, and -8. This is due to their proposed function in the regulation of TG metabolism. These ANGPTLs are proposed to act as inhibitors of TG clearance, leading to accumulation of TG in circulation with increases in these proteins. ²² After 5 days of CSO diet enrichment, there were greater decreases in fasting ANGPTL-4 and -8 compared with OO diets. ²³ Further, in the same study, there were greater decreases in postprandial ANGPTL-3 and -4 after CSO versus OO enriched diets. ²³ Conversely, in adults with hypercholesterolemia, 8-weeks of CSO enrichment did not change ANGPTL responses, although, there were increases in ANGPTL-3 and -4 postprandial responses after the OO diet. ²⁴ These ANGPTL changes correspond to the observed changes in fasting and postprandial TG observed in the same cohorts, ^10,11 suggesting the influence of ANGPTLs from CSO consumption may be a mechanism through which it offers improvement or protection of TG metabolism in a positive energy balance state. Postprandial glucose was only measured in the study involving adults with hypercholesterolemia. Eight weeks of CSO consumption resulted in non-significant suppression of postprandial glucose; however, this was significantly improved compared with increased postprandial glucose after OO consumption. ¹¹ These changes in glycemic response somewhat mirror the effects observed for postprandial TGs and ANGPTLs suggesting CSO may have an overall protective effect on postprandial metabolism, potentially mediated through ANGPTLs. For coagulation potential, two markers have been measured in response to CSO diet enrichment: plasminogen activator inhibitor-1 (PAI-1), and tissue factor (TF). CSO has been shown to lower TF in young males ¹⁶ with no changes in adults with hypercholesterolemia. ¹⁷ This last result further emphasizes the potential cardiovascular improvements associated with CSO consumption, depending on the population studied.

CONCLUSIONS, LIMITATIONS, AND FUTURE WORK

Both the animal model work and human clinical trials show the potential of CSO to support cardiometabolic health when incorporated into the daily diet. CSO has been shown to benefit fasting and postprandial blood lipids and ANGPTLs, improve measures of physiological appetite, and in one instance, reduce self-reported energy intake. These findings are in the context of HF diet consumption and showed no evidence of increasing markers of inflammation or coagulation potential.

While the literature to date shows promising evidence for the potential of CSO to serve as a viable nutrition strategy for supporting cardiometabolic health, there are many gaps left to be filled in. First, the dose of CSO required to elicit these cardiometabolic improvements has never been evaluated. Estimation of an appropriate dose has been attempted by Hart et al. ²⁵ however, this was based on two trials both utilizing HF diets. This also emphasizes that, to date, only two randomized clinical trials have been conducted on CSO diet enrichment further limiting this data. Moreover, CSO has only been compared with OO in clinical trials and has never been compared in humans to other PUFA-rich oils. In animal models, CSO has been shown to have superior lipid lowering effects compared with corn, ^{6 –8} soybean, ¹⁴ and safflower oil ¹² diets, all of which are high in PUFAs, but this effect has not been confirmed in humans. Additionally, CSO has never been investigated for periods longer than 8 weeks, or in the context of other metabolic conditions such as type 2 diabetes mellitus or metabolic syndrome. Lastly, there are many other clinical outcomes that have never been assessed after CSO consumption, some of which include hepatic function, oxidative stress, and vascular function. The presented limitations of the existing CSO literature point to some exciting future directions. In conclusion, CSO may be a promising option to potentially serve as a simple dietary approach to preserve and/or improve cardiometabolic health.