Dietary strategies that reduce post-prandial glycemia are important in the prevention and treatment of diabetes and coronary heart disease (CHD). This may be achieved by addition of high-quality protein and fat contained in pistachio nuts, to carbohydrate-containing foods or meals.
A total of 10 healthy volunteers (3 males, 7 females); aged 48.3±6.4 years; Body mass index (BMI) 28.0±4.8 kg/m2 participated in two studies. Study 1 assessed the dose-response effect of 28, 56 and 84 g pistachios consumed alone or co-ingested with white bread (50 g available carbohydrate); Study 2 assessed the effective dose (56 g) of pistachios on post-prandial glycemia consumed with different commonly consumed carbohydrate foods (50 g available carbohydrate). Relative glycemic responses (RGRs) of study meals compared with white bread, were assessed over the 2 h post-prandial period.
The RGRs of pistachios consumed alone expressed as a percentage of white bread (100%) were: 28 g (5.7±1.8%); 56 g (3.8±1.8%); 84 g (9.3±3.2%), P<0.001. Adding pistachios to white bread resulted in a dose-dependent reduction in the RGR of the composite meal; 28 g (89.1±6.0, P=0.100); 56 g (67.3±9.8, P=0.009); 84 g (51.5±7.5, P<0.001). Addition of 56 g pistachios to carbohydrate foods significantly reduced the RGR: parboiled rice (72.5±6.0) versus rice and pistachios (58.7±5.1) (P=0.031); pasta (94.8±11.4) versus pasta and pistachios (56.4±5.0) (P=0.025); whereas for mashed potatoes (109.0±6.6) versus potatoes and pistachios, (87.4±8.0) (P=0.063) the results approached significance.
Pistachios consumed alone had a minimal effect on post-prandial glycemia and when taken with a carbohydrate meal attenuated the RGR. The beneficial effects of pistachios on post-prandial glycemia could, therefore, be part of the mechanism by which nuts reduce the risk of diabetes and CHD.
Post-meal hyperglycemia is an independent risk factor for cardiovascular disease (CVD) (Levitan et al., 2004) and a condition that is common in patients with type 2 diabetes (Akbar, 2003; Bonora et al., 2006). Dietary glycemic index (GI) and glycemic load have been shown to reliably predict the relative post-meal glycemic and insulinemic responses to mixed meals (McMillan-Price et al., 2006; Wolever et al., 2006), and evidence suggests that lowering the GI of a diet can positively affect post-meal plasma glucose excursions (Opperman et al., 2004) and as such, potentially reduce the risk of CVD and diabetes (Salmeron et al., 1997a, 1997b; Liu et al., 2000)
Epidemiological and interventional studies have linked the intake of tree nuts to reduced risk of CVD and diabetes (Jiang et al., 2002; Li et al., 2009; Mente et al., 2009). This protective effect may in part be explained by the favorable impact of nuts on serum lipids (Griel and Kris-Etherton, 2006), oxidative stress (Jenkins et al., 2002) and markers of inflammation (Jiang et al., 2006; Ros, 2009). However, owing to their low available carbohydrate content and favorable fat and protein profiles, nuts may also decrease the risk of CVD and diabetes by reducing post-prandial blood glucose excursions. This mechanism has only been examined in a few clinical studies, which have shown that almonds eaten alone have a minimal effect on blood glucose, and when different doses are consumed with a carbohydrate meal, the post-prandial glucose response is attenuated (Josse et al., 2007).
To further test the potential impact of nuts on post-meal hyperglycemia, we conducted two studies using pistachio nuts. The first assessed the dose-response effect of pistachios on post-prandial blood glucose when consumed alone, and consumed with a white bread meal. The second study assessed the effect of pistachios on post-prandial blood glucose excursions when added to other commonly consumed carbohydrate-rich foods.
Subjects and methods
A total of 10 overweight but otherwise healthy subjects (3 males and 7 females), aged (mean±s.d.) 48.3±6.4 years with a body mass index of 28.0±4.8 kg/m2, were studied. Subjects were not taking any medications that would interfere with glucose metabolism.
Subjects were recruited through local advertisement and through the clinic volunteer roster. Each subject underwent treatments on separate days, performing up to two tests per week with at least one day in between tests. On each test day, subjects came to the research clinic in the morning after a 10–14 h overnight fast. After being weighed and having a fasting blood sample obtained by finger-prick using autolancets (Microlet, Bayer, Toronto, Canada), the subject then consumed a study meal within 10 min, and further finger-prick blood samples were obtained at 15, 30, 45, 60, 90 and 120 min after the start of the meal. Subjects were given a choice of one or two cups of water, tea or coffee with or without milk. The beverage consumed by each subject remained the same for each of their study days. Subjects were instructed not to consume nuts the day before each study day.
The study protocol was approved by the Western Institutional Review Board. Written informed consent was obtained from all subjects before the start of the study.
Meals and palatability
The tests meals in study 1 consisted of three different doses (28, 56 and 84g) of pistachio nuts either consumed alone or together with white bread (50 g available carbohydrate) (Table 1). In study 2, 50 g available carbohydrate portions of rice, pasta or instant mashed potatoes were consumed with or without 56 g of pistachios (Table 1). The control white bread was also consumed on three different occasions to allow calculation of the relative glycemic response (RGR). The control white bread was baked in a bread maker in loaves containing 250 g available carbohydrate. The ingredients for each loaf (250 ml warm water, 340 g all-purpose flour, 7 g sugar, 4 g salt and 6.5 g dry yeast) were placed into the bread maker according to instructions, and the machine was turned on. After the loaf was prepared, it was allowed to cool for an hour, and then weighed and after discarding the crust ends, the remainder was divided into portion sizes containing 50 g available carbohydrate. These portions were frozen before use, and reheated in a microwave before consumption. Study meals were given in random order.
Palatability was rated on a 100 mm visual analog scale anchored with ‘very unpalatable’ at one end and ‘very palatable’ at the other. The higher the number, the higher was the perceived palatability of the product.
Blood glucose analysis
Blood samples were stored at −20 °C until analysis, which took place within 4 days of collection. Blood glucose was analyzed using a Yellow Springs Instrument Model 2300 STAT (Yellow Springs, OH, USA).
Incremental areas under the plasma glucose curves were calculated using the trapezoid rule, ignoring area beneath the baseline value. The GI or RGR was calculated by expressing each subject's glucose incremental areas under the plasma glucose curves for the test food as a percentage of the same subject's average response after consuming the reference white bread. The mean of the resulting values was the RGR. To calculate the RGR, based on the glucose scale (that is, GI of glucose=100), the RGR bread scale values were multiplied by 0.71. The blood glucose concentrations at each time point and the incremental areas under the plasma glucose curve values were subjected to repeated-measures analysis of variance, examining for the effect of test meal. After demonstration of significant heterogeneity, the significance of the differences between individual means was assessed using Tukey's test to adjust for multiple comparisons. In addition, the significance of the differences between blood glucose concentrations and increments for each test food and white bread were assessed by paired t-test. To estimate the RGR lowering potential of 1 g of pistachios, the change in RGR for each study meal was averaged and calculated per gram of pistachio consumed.
Relative glycemic response
Mean fasting blood glucose was identical before each test meal within each series. The glycemic response to all three doses of pistachios was significantly lower at 30, 45, 60 and 90 min after ingestion in comparison with the white bread control (P<0.01) (Figure 1a). Furthermore, 28, 56 and 84 g of pistachios had significantly lower RGR in comparison with white bread (5.7±1.8, 3.8±1.8 and 9.3±3.2 versus 100, respectively; P<0.001) (Table 2). Similarly, adding increasing doses of pistachios to white bread significantly reduced the glycemic response at 30, 45, 60 and 90 min for the 84 g dose (P<0.01), at 45, 60 and 90 min for the 56 g dose (P<0.01) and at 45 and 60 min for the 28 g dose (P<0.01) in comparison with the white bread control (Figure 1b). When pistachio nuts were added to the white bread meal, only the higher doses, 56 and 84 g, resulted in significant reductions in the RGR in comparison with the white bread control (67.3±9.8 (P=0.009) and 51.5±7.5 (P<0.001) versus 100, respectively) (Table 2).
In study 2, the addition of 56 g of pistachios to other commonly consumed carbohydrate rich foods (parboiled rice, pasta and potatoes) resulted in reduced glycemic responses (Table 3). When the pistachio meals were compared with their respective control meals, the addition of 56 g of pistachios significantly lowered the RGR of both parboiled rice (72.5±6.0 versus 58.7±5.1; P=0.031) and pasta (94.8±11.4 versus 56.4±5.0; P=0.025), whereas the reduction approached significance for instant mashed potatoes (109.0±6.6 versus 87.4±8.0; P=0.063) (Table 3; Figure 2). The mean reduction in glycemic response observed per gram of pistachios consumed was found to be 0.35 U, that is, 1 g of pistachios added to a food lowered the RGR of that food by 0.35 units on the glucose scale or 0.49 U on the bread scale. It must be pointed out that these results may only apply to the foods tested and results may be different for foods varying in composition and physical form (for example, beverages).
The present study is the first to assess the effect of pistachio nuts on post-prandial glycemia. The findings confirm the very low glycemic responses to pistachios eaten alone would be predicted on the basis of their low available carbohydrate content (6 g per ounce) and on the results of other nuts (almonds) and oil seeds (peanuts), (Johnston and Buller, 2005; Jenkins et al., 2006; Josse et al., 2007). Moreover, these data demonstrate that the addition of pistachios to foods with high available carbohydrate content reduces the overall glycemic impact of the foods studied (parboiled rice, pasta, white bread and mashed potatoes), despite increasing the overall available carbohydrate content. Of particular interest was the progressive reduction in bread glycemic response with the increasing doses of pistachios. The graded glycemic response reduction of white bread has also been observed with almonds in a similar dose-response study conducted in healthy individuals (Josse et al., 2007). Pistachios and other nuts are energy-dense foods, and it has been recognized that gastric emptying is reduced by a high fat and energy load such that the pylorus tends to regulate the flow of energy into the duodenum (Hunt and Stubbs, 1975; Peracchi et al., 2000). A study by Henry et al. (2008) found that the glycemic response of bread can be lowered by the addition of any type of fat. Thus, increasing energy and fat load with an increasing dose of pistachios would therefore tend to reduce the glycemic response in a dose-dependent manner as observed. It is also possible that the addition of pistachios resulted in a reduction in carbohydrate absorption or altered the osmotic load and volume of the stomach, all of which would alter the post-prandial glycemic response. The acute benefits of pistachios on glycemia observed in the current study may be indicative of long-term improvements in glycemic control in light of recent evidence showing that the addition of 60–100 g of pistachios to a Mediterranean diet for 4 weeks significantly reduced fasting glucose levels in healthy adults in comparison with an unmodified Mediterranean diet (Sari et al., 2010).
The low post-meal glycemic response to pistachios is also of particular interest in relation to recent studies, indicating that nut consumption was protective for CVD (Fraser et al., 1992; Hu et al., 1998; Ellsworth et al., 2001; Kris-Etherton et al., 2001; Albert et al., 2002). In each of these large prospective studies, the Physicians’ Health Study (Albert et al., 2002), the Iowa Women's Health Study (Ellsworth et al., 2001), the Nurses’ Health Study (Hu et al., 1998) and the Adventist Health Study (Fraser et al., 1992), those who consumed the greatest amount of nuts versus those who consumed the least had significant reductions in the risk of death from CVD, relative risk=0.53, 0.81, 0.65 and 0.52, respectively. Clinically, post-meal hyperglycemia has been linked to endothelial dysfunction (Williams et al., 1998; Kawano et al., 1999) and oxidative stress (Ceriello et al., 2004); both of which are CVD risk factors. Post-prandial lipemia has also been associated with an increased risk for coronary heart disease (CHD). Although consumption of nuts would increase post-prandial lipemia, a study by Berry et al. (2008) found that eating whole almonds resulted in a lower post-prandial rise in plasma triglycerides compared with almond oil or sunflower oil meals (74 and 58%, respectively) that were balanced for macronutrients and fiber. In addition, nuts have a healthy fatty acid profile that is largely unsaturated and may provide other CHD benefits beyond reducing the post-prandial glycemic impact of a meal. Monounsaturated fat tends to raise high-density lipoprotein cholesterol when exchanged for carbohydrate, and the Mensink and Katan equation also attribute a small but important low-density lipoprotein cholesterol lowering effect to monounsaturated fat (Mensink and Katan, 1992). This lowering of low-density lipoprotein cholesterol is supported by a number of studies, which have shown that pistachios (Edwards et al., 1999; Kocyigit et al., 2006; Sheridan et al., 2007; Gebauer et al., 2008; Kay et al., 2010; Sari et al., 2010), almonds (Griel and Kris-Etherton, 2006) and walnuts (Banel and Hu, 2009) can potentially reduce the risk of CVD by improving serum lipid concentrations.
The low glycemic effect of nuts may provide a reason for the inclusion of nuts in diets aimed at reducing the risk of diabetes as the substitution of nuts for high GI carbohydrates would reduce the glycemic load of the diet. Furthermore, it may be beneficial to advise people with Type 2 diabetes to consume nuts when they consume carbohydrates as part of their daily diet to reduce the acute, post-prandial glycemic impact of the meal. These recommendations are supported by recent recommendations by both the American Diabetes Association and the International Diabetes Federation, which emphasize management of post-meal hyperglycemia as an important component in the management and prevention of Type 2 diabetes (American Diabetes Association (ADA), 2007; International Diabetes Federation, 2007).
Few studies have assessed the effect of nuts on diabetes risk and control. Of the two cohort studies, only the results from the Nurses Health Study cohort (Jiang et al., 2002) but not the Iowa Women's Health Study (Parker et al., 2003) indicate that consumption of five or more servings per week of nuts compared with rarely or no intake reduces the risk of developing diabetes (relative risk=0.73, 95% confidence interval, 0.60–0.89). Data from clinical studies indicate that nuts are beneficial in reducing CHD lipid risk factors in patients with diabetes or the metabolic syndrome (Lovejoy et al., 2002; Tapsell et al., 2004; Estruch et al., 2006; Li et al., 2009). In addition, some studies have found a significant reduction in fasting insulin (Casas-Agustench et al., 2011; Tapsell et al. 2009) or a significant improvement in insulin resistance (Casas-Agustench et al., 2011). However, none have reported a significant reduction in glycated proteins as a marker of long-term glycemic control (Lovejoy et al., 2002; Scott et al., 2003; Tapsell et al., 2004; Estruch et al., 2006; Mukuddem-Petersen et al., 2007; Casas-Agustench et al., 2011; Tapsell et al., 2009; Ma et al., 2010). In light of recent evidence showing that the combination of the 2 h post-meal glucose levels and fasting blood glucose allow for better estimation of risk both for diabetes and CVD than the fasting blood glucose alone (Sorkin et al., 2005; International Diabetes Federation, 2007); further studies that examine the effect of nuts on post-meal glycemia are warranted.
We conclude that pistachios resulted in a minimal rise in blood glucose when fed alone over a wide range of acceptable intake levels. Moreover, they also depress the blood glucose response to bread in a dose-dependent manner and the glycemic response of other carbohydrate rich meals, possibly related to their high fat, fiber and protein content. In view of their favorable fatty acid profile, their beneficial effects on serum lipid levels and their low glycemic impact, it seems now appropriate to assess their effects in longer-term studies of diabetes. These studies should aim at reducing the dietary glycemic load to assess whether nut consumption results in reduced concentrations of long-term biomarkers of glycemic control, such as glycated proteins. However, it is important to note that in the context of a healthy diet, care must be taken not to increase calories above requirement and nuts should be substituted for other foods and not simply added to one's diet.
We are grateful to the Western Pistachio Association, Fresno, CA, USA for provision of the pistachios and for funding for additional statistics. This work was supported by the Western Pistachio Association, Fresno, CA, USA and the California Pistachio Commission. DJAJ is funded by the Federal Government of Canada as a Canada Research Chair in Nutrition and Metabolism.
About this article
Consumption of fatty foods and incident type 2 diabetes in populations from eight European countries
European Journal of Clinical Nutrition (2015)