Originally Published:20040701.

Introduction

Public health nutritionists have historically recommended variety or diversity in dietary patterns as one means of fostering an optimal diet. The underlying concept was that no one food contained all of the necessary nutrients and that variety in dietary sources was needed to ensure a "balanced" diet.

The issue of diversity or variety in the diet has recently become more controversial. The present paper reviews the issue of dietary diversity in relationship to:

* diet quality

* overweight and obesity

* food security

Historical Perspective

Historically, diet diversity or variety has been an essential element of food-based dietary guidelines. A recent report by the International Task Force on Nutrition and Optimal Development1 reviewed food-based dietary guidelines worldwide and concluded that, "The focus of food-based dietary guidelines is on a balanced and moderate intake of a variety of nutritious foods." Indeed, 25 countries, including Mexico, have a variety guideline as part of their national dietary guidelines (Table 1). Japan has the most quantitative variety guidelines, which recommend consuming more than 30 different kinds of foods each day. Essentially, countries with food-based guidelines have placed an emphasis on balance, moderation and variety.

In reviewing the individual country dietary guidelines, there is a clear assumption that dietary variety is a well-understood and accepted concept. However, there are clear differences throughout the scientific literature in how dietary variety is measured.

Dietary Variety and Diet Quality

In the past decade, several researchers have attempted to develop a measure of diet quality, both with and without a variety component.

Kennedy2 developed a single summary index of diet quality called the Healthy Eating Index (HEI). The HEI is an index ranging from zero to 100, which is based on ten individual components. As can be seen from Table 2, the individual component scores can vary from zero to ten. The first five components of the HEI are based on the five major food groups of the US Food Guide Pyramid,3 while components six to ten are based on aspects of the Dietary Guidelines.4 Variety is one of the 10 components of the HEI. Results from a nationally representative sample of Americans2 indicated that the HEI:

* correlated positively and significantly with most nutrients in the diet

* correlated with the body mass index (BMI, kg/m2) of study subjects

* correlated significantly with the individuals "self perception" of their diets. People who rated their diets as poor or fair had a lower HEI than individuals who rated their diets as good or excellent.

No one component of the index dominated the HEI score. In developing the index there was a need for ways to quantify dietary variety, but the existing published literature contained little guidance on how to translate consumption of existing foods into a quantification for variety. The HEI gave a perfect score of 10 on the variety component if the individual consumed 16 or more different foods over a 3-day period; a score of zero was given if 6 or fewer different food items were consumed over 3 days. Foods were counted as part of the variety score only if they were eaten in sufficient quantity to contribute at least one half a serving in one of the food groups. Identical items eaten on separate occasions were aggregated before imposing the one-half serving cutoff point. The variety component contributed independently to the overall HEI score. People with more variety in their consumption patterns had a significantly better diet quality.

Drewnowski and colleagues5 took a different approach in quantifying dietary variety. The Dietary Variety Score (DVS) was based on the number of different foods across a 15-day period. The results indicate that a higher DVS was positively, significantly associated with vitamin C consumption and negatively significantly correlated with salt, sugar and saturated fat. Drewnowski et al. also developed a Diet Quality Score (DQS) based on conformity to the US Dietary Guidelines for total fat, saturated fat, cholesterol, sodium, and carbohydrate. Results from this study indicated that the DVS- was not linked significantly to the Diet Quality Score.

Dietary Variety, Overweight, and Obesity

The US Dietary Guidelines for Americans were relatively unchanged from 1980 to 19954; each revision of the Dietary Guidelines up until 1995 continued to maintain seven core guidelines, with variety either as the first guideline or as in 1995, the central guideline around which the other guidelines revolve. The wording for the guideline "Eat a Variety of Foods" was unchanged from 1980 to 1995. The importance of the variety guideline stemmed from nutritional adequacy. There was a known association between variety and achieving the RDAs for individual nutrients; this same relationship, however, was not observed between variety and a range of macronutrients, including total fat, saturated fat, and sodium.

The 2000 U.S. Dietary Guidelines broke with tradition and changed the variety guideline to "Let the Pyramid Guide Your Food Choices." The reasons for this change were threefold:

* The Expert Dietary Guidelines Advisory Committee Report6 noted that the Food Guide Pyramid was an integral part of achieving dietary variety. The report also indicated that there was no evidence that choosing more foods within a food group led to a "better overall diet."

* Evidence from consumer research suggested that the earlier variety guidelines might not be clear to consumers. Focus group research6 suggested that some individuals saw the variety guidelines as a license to consume foods that were not healthful choices.

* Whereas consumers know and understand the Food Guide Pyramid, there was concern that eating a variety of foods might promote overconsumption of energy.

This last point was influenced by research.7 Detailed metabolic research on a small group of adults in Boston indicated that variety within the grain group of the Pyramid was associated with increased energy consumption. However, variety within the vegetable group had a significant, negative association with energy intake. The authors were clear to emphasize that it may not be variety per se, but variety in only selected foods groups that is linked to a higher risk of overweight and obesity. Interestingly, while the variety wording was replaced by reference to the Pyramid in the 2000 Dietary Guidelines for Americans, the word "variety" continues to be used in the other guidelines. For example, two of the guidelines suggest that people "choose a variety of grains daily, especially whole grains" and "choose a variety of fruits and vegetables daily."

It is not clear why variety is useful in some, but not all, advice to consumers. A recent review of 39 animal and human studies8 came to the overall conclusion that "Food consumption increases when there is more variety in a meal or diet; greater variety is positively associated with increased body weight and body fat." The authors go on to conclude that according to available research, a reduction in dietary variety of highly palatable, energy-dense foods may be useful in the treatment and prevention of obesity.

A 2001 study9 came to somewhat different conclusions on the variety/overweight links. Using a nationally representative sample of more than 10,000 adults ages 19 and older, the study was designed to assess the effects of different types of popular diets on (1) diet quality, (2) macronutrient consumption, and (3) BMI. The diets with greater than 55% of energy from carbohydrate contained more variety and were also significantly more likely to have lower energy and a lower percentage of kcals contributed by total fat. The individuals eating the high-carbohydrate diet also consumed a significantly less energy dense diet when compared with the low-carbohydrate category. In addition, the average BMIs for men and women were significantly lower for people consuming the high carbohydrate diet with more variety and this group was significantly less likely to have subjects with a BMI greater than 25.

The same study9 further categorized the high-carbohydrate group into those people eating versus those not eating according to the Food Guide Pyramid. The diet quality as measured by the HEI was significantly higher for those in the Pyramid group compared with those in the non-Pyramid subjects. The energy intake, however, was significantly higher for subjects in the Pyramid group but the BMI was significantly lower for men and also lower for women following a Pyramid-type diet. In this study, more variety was associated with a better quality diet and a BMI more likely to be in the normal range.

These disparate results are not as contradictory as they might first seem. The McCrory7 research emphasized that increased variety, primarily in the grain group, was associated with the increased overweight and obesity. In the Raynor and Epstein8 review, the variety of energy-dense foods was highlighted as increasing the risk of overweight and obesity. In the research by Kennedy et al.9 variety was associated with more fruit and vegetable consumption and in most groups, a lower energy consumption and thus, a lower BMI.

Dietary Variety and Food Security

Food security is the foundation of a healthful diet. It has been well documented that at low levels of income, households and individuals try to diversify their dietary patterns.10 Worldwide, as we see populations in transition from rural to urban areas, from low income to higher income, differences in dietary patterns are occurring.11 In general, grain-based diets are being replaced by diets with more variety including more total fat, saturated fat, and sugar. This is true at all income levels. Therefore, the issue for this paper is how dietary variety is related to food security.

A 10-country study12 was conducted using data from Ghana, Malawi, Mali, Kenya, India, the Philippines, Mozambique, Mexico, Bangladesh and Egypt. The research explored the relationship between number of unique foods consumed (variety), household food expenditures per capita, and household energy availability per capita. Results indicate that in each of these 10 countries, there was a positive, significant association between diet diversity and household per capita expenditures. In addition, in all countries, there was also a positive, significant association between household diet diversity and household calorie availability per capita. This was true for both urban and rural areas.

A recent review for FAO on newer measures of food insecurity noted that households with hungry adults and hungry preschool-aged children lacked both the quantity and quality of food for a healthful diet.13 The apparent global paradox of obesity that occurs simultaneously alongside food insecurity is related, in part, to the cost of an optimum diet. Households with limited incomes may be trying to maximize energy per dollar spent. This results in eating patterns that are excessive in total fat and saturated fat, as well as limited variety.14