Raisins are fruit, but they’re also about 79% carbohydrate by weight. Yet, according to this study, participants who consumed raisins had a significant 23% reduction in postprandial glucose levels as compared to those who consumed a 100-calorie “snack pack.”
Bays H, Weiter K, Anderson J. A randomized study of raisins versus alternative snacks on glycemic control and other cardiovascular risk factors in patients with type 2 diabetes mellitus. Phys Sportsmed. 2015;43(1):37-43.
This 12-week randomized efficacy study evaluated the impact of routine consumption of dark raisins vs alternative processed snacks on glucose levels and other cardiovascular risk factors among patients with type 2 diabetes mellitus (DM-2).
Fifty-one participants with DM-2
Study Medication and Dosage
Participants in the raisin treatment arm consumed a 1-oz package (90 calories) of raisins 3 times a day before meals with a glass of water or a noncaloric beverage (eg, tea). Participants randomized to the snack group consumed a 100-calorie snack pack 3 times a day before meals with water or another noncaloric beverage.
Postprandial glucose level, glycated hemoglobin (HbA1c) level, blood pressure (BP), and cardiovascular risk factors including weight, waist circumference, and cholesterol levels
Study participants who consumed raisins had a significant 23% reduction in postprandial glucose levels (P=0.024). Also in comparison to study participants who ate packaged snacks, those who consumed raisins had a 19% reduction in fasting glucose and a 0.12% reduction in HbA1c, although these latter findings did not achieve statistical significance. Regarding BP, those who consumed raisins had a significant 8.7 mmHg reduction in systolic BP (P=0.035) compared to the 100-calorie snack group but did not experience a significant reduction in diastolic BP. Compared to the snack group, the raisin group did not have a significant improvement in body weight, body mass index, waist circumference, fasting insulin, homeostatic model assessment of insulin resistance, total cholesterol, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, triglyceride, or non-HDL cholesterol levels.
It has become common practice among food manufacturers to package what we might describe as junk food in 100-calorie, single-portion packages. These packs are marketed as a way to help dieters control their caloric intake while still satisfying the urge for their favorite snack foods. Food manufacturers love this concept—and apparently, the public does as well. In 2007, The New York Times reported that in just the 3 preceding years, the “snack pack” portion of the food market had grown by 200 million dollars per year in sales. The author of the article noted:
As a business concept, the idea is simple. Take an existing product, portion smaller amounts of it into single-serving bags, and sell several of the bags for about the same or more as a regular-size package. Consumers do not seem to mind paying more even though they are getting fewer Goldfish.1
The Times article quoted 1 food industry expert as saying, “The irony is if you take Wheat Thins or Goldfish, buy a large-size box, count out the items and put them in a Ziploc bag, you’d have essentially the same product” but at a 20% lower cost.1 Food company executives readily admit the smaller packages do cost more but that they make consumers “exercise a little hand-to-mouth restraint.”
The greatest implication of this study is perhaps this reminder that the American public will succumb to anything that promises weight loss, even if it costs a great deal more than simple self-moderation.
The fact that 40% of consumers will pay 20% to 30% more for convenience says something about the American consumer that may be more important to us as practitioners than these data on raisins. The greatest implication of this study is perhaps this reminder that the American public will succumb to anything that promises weight loss, even if it costs a great deal more than simple self-moderation.
The current study by Bays et al found that eating raisins is better for our health than eating these snack packs. Is that a surprise? If we look closely, it is. Raisins are pretty much straight sugar: 32.5 g raisins contain 25.7 g carbohydrates and 5 g water. Of those carbohydrates, 19.6 g are sugars, 9 g glucose, and 9.7 g fructose. Raisins are thus about 79% carbohydrate by weight with a glycemic index of about 66.2
The website CalorieCount.com provides links to nutritional information of 100 different snack packs
, and the list isn’t getting smaller with time. These snack packs, while all adding up to about 100 calories, vary in carbohydrate, protein, and fat content. It’s interesting that the nutritionists at Calorie Count give Blue Diamond plain almonds a grade of B‒, seemingly for exceeding the calorie count: a 1-oz serving contains 160 calories primarily from fat. On the other hand, they award Hostess Miniature Cinnamon Coffee Cakes a solid B, perhaps because 3 of these tiny items account for only 100 calories. Of the 100 calories in these coffee cakes, though, 5% come from fat, 84% from carbohydrates (24% from sugar), and 8% from protein. These snacks in general set the bar low for intelligent nutritional choices, but they are technically not that all different than a similar serving of raisins. Keep in mind that the Prevención con Dieta Mediterránea trial did report that a daily serving of nuts, akin to the almond snack pack, is associated with a 30% drop in cardiovascular events in a high-risk population3,4
; the scoring system on this calorie count website does not appear to be based on concerns about health but simply calories.
Still, raisins appear to be a healthier snack than the commonly chosen alternatives.
There is a bigger issue here, one that is worth pondering. It is our tendency to think that a healthy diet can be summarized by total calorie intake or for that matter, carbohydrate, fat, and protein intake or even glycemic impact. Biological actions are not that simple.
People—and this includes naturopathic doctors—tend to simplify life’s complexities and frequently attempt to group related items into manageable categories, to sort things out so they make sense. Life (and in particular biology) is complicated, and our habit of simplification often leads us to inaccuracies in thinking and inaccuracies in predicting long-term health impact of consuming specific foods. This study on raisins is a reminder.
There are amazingly large numbers of edible foods in the world. A common estimate is that there are about 20,000 species of edible plants (although just 20 species provide 90% of our food supply).5 Modern dieticians have conveniently simplified their business by creating categories to group these foods into: fruits, vegetables, starches, oils, meats, and types of fish. Our global cornucopia has been further simplified, often referred to in terms of carbohydrates, protein, and fats. The assumption is that all members of a category will have similar if not identical nutritional impact when they are consumed. While sometimes this kind of thinking works, here it doesn’t.
This study is one of several suggesting the concept of “fruit” may need alteration and that considering various fruits as interchangeable may be an oversimplification. This inadequacy surfaced with Muraki et al’s 20136
paper on risk of diabetes and fruit consumption, a paper that one of this article's authors (Klearman) reviewed for this journal
Muraki’s group used data from 3 large cohorts—the Nurses’ Health Study I and II and the Health Professionals Follow-up study—comprising a total of 187,382 people followed over decades. Muraki estimated the effect of consuming individual fruits or fruit groups (that were grouped together by glycemic index and load) and fruit juice in relation to the risk of developing DM-2. Their data suggest that consuming whole fruit lowers the risk of developing DM-2. For every 3 servings of fruit consumed per week, the risk of developing diabetes decreased by 2%.
Certain fruits had greater or lesser impact on overall diabetes risk. When adjusting for individual fruits, for every 3 servings per week, blueberries decreased risk by 26%, grapes and raisins by 12%, prunes by 11%, apples and pears by 7%, and bananas by 5%. Grapefruit did not have a statistically significant impact on DM-2 risk, but cantaloupe increased risk by 10%. Our desire for simple explanations for why these differences occurred was thwarted; the glycemic index/glycemic load values of each fruit did not fully account for these differences in disease risk. The impact of fruit on health is more complicated than we thought.
It also appears that the effect of whole fruit is different than that of fruit juice. Muraki reported that drinking 3 servings of fruit juice each week increased the risk of developing DM-2 by 8%. While O’Conner et al in a May 2015 study did not confirm this association between fruit juice and diabetes risk, they did confirm that, in general, greater consumption of sweet beverages is associated with higher risk for diabetes.8
Given the public’s current widespread enthusiasm for fruit smoothies, this difference in risk between whole fruit and fruit juice, if confirmed, could have significant clinical implications. This difference in long-term effect could be, as Klearman pointed out in her earlier commentary, because fruit juice is devoid of the soluble fiber that helps to decrease the absorption of glucose into the bloodstream.9 Absence of fiber in fruit juice leads to faster and greater changes to serum glucose and insulin levels.10 However, Muraki et al suggested that the difference in risk from whole fruit has nothing to do with differences in glycemic impact; rather, it more likely results from the different types and quantities of phytochemicals contained in each fruit. Anthocyanins reduce the amount of glucose produced in the liver.11 We should consider resveratrol, chlorogenic acid, and naringin levels of individual fruits as well. The amount and variety of these various compounds is not distributed equally in all fruits. We would prefer simpler explanations, but simple explanations just may not be accurate.
This current study by Bays et al on raisin consumption expands on this idea. Results of trials in which participants snacked on raisins published in 2014 are of interest. One study reported that snacking on raisins 3 times a day was associated with lower postprandial glucose levels (P=0.003 vs baseline; P=0.03 vs snacks) and that eating raisins appeared to significantly decrease HbA1c levels (‒0.12%; P=0.004), a significantly greater decrease than seen with other equal calorie snack intakes. Eating raisins was also associated with significant reductions in both systolic and diastolic BP.12 Another 2014 study in which diabetic participants ate 36 g raisins only twice a day found that this intervention was associated with significant decreases in diastolic BP but not in changes in glycemic control.13
Raisins appear to have a lower impact on blood sugar than one would predict. In 2013 Kanellos and colleagues reported that they had given both diabetic and healthy subjects either raisins or an equivalent amount of glucose and then tracked changes in blood sugar at short intervals as in a glucose tolerance test. Neither blood sugar nor insulin rose as much after the raisins as it did after the glucose.2 Another 2013 clinical study reported raisin consumption was associated with increased satiety and decreased food consumption, as well as significant reductions in LDL cholesterol, triglycerides, and oxidized LDL cholesterol levels.14
In another study, researchers found that giving children all the raisins they want as an afterschool snack lowers their total daily calorie intake compared to consumption of potato chips or cookies.15 The reason raisins are so filling may be that they are very high in fiber, containing about 5 g per 100 g soluble and insoluble fibers. In addition the same 100 g of raisins contain about 5 g fructooligosaccharides that form during drying and are not present in whole grapes. So raisins actually contain about 10 g fiber per 100 g or about 10% of their weight.16 As such, raisins are an easy and relatively palatable way for Americans to increase fiber intake.17
In vitro work suggests that the phenolic compounds in raisins may be protective against colon cancer development. Raisin extracts suppressed colon cancer cell proliferation, decreased glutathione (which can protect cancer cells), and also decreased cyclooxygenase 2, potentially decreasing chronic inflammation.18 Thus the impact of raisins, at least on risk of diabetes, is different from that of other fruits and possibly even different from grapes. The sugar content and glycemic index of fruits, while tempting as an explanation for increasing risk of diabetes, are inadequate.
This matter brings to mind a previous discussion on the work of Törrönen
on berries and their effect on the glycemic index of bread written by one of the authors of this article (Schor).20
Törrönen reported that eating strawberries, bilberries, lingonberries, or chokeberries with white bread significantly reduced the postprandial insulin response, and so did eating a mixed berry puree with either white bread or rye bread. Both strawberries and the mixed-berry puree significantly improved the glycemic profile of both breads (about 40% for the white bread and about 20% for the rye). When white bread is consumed with berries, less insulin is needed for maintenance of a normal or even slightly improved postprandial glucose metabolism. Even the insulin profile of rye bread, which is initially better than that of white bread, improved. Few of us would guess that adding a berry puree to bread lowers impact on blood sugar. It makes sense only when we remember the evolutionary role of berries is to attract birds that consume them and then increase seed dispersal when the laxative impact of the berries takes action. For this process to be effective, it is best if the berries taste good while they interfere with digestion. This is another argument against lumping all fruits together and certainly against the habit of thinking of them as just another source of sugar.
As for the present study, these are the takeaways. Raisins make good snacks in comparison to snack packs. Different fruits have varying impacts on health, and we should be cautious about generalizing. Finally we should ponder whether fruit smoothies will act more like whole fruit or fruit juice when it comes to risk of diabetes. If the later, the public's current enthusiasm for consuming fruit smoothie drinks may eventually come to be associated with increased risk of diabetes.