Introduction

As a society we have become more aware of the implications of our food choices with regards to our general health. We are taught by media and health professionals to think before we eat and to make "healthy food choices". In many ways, watching what we eat has become an important step in achieving and maintaining a healthy lifestyle. For many individuals, however, the choice of what and how much to eat is an almost constant stress. Extreme dieting behavior and the prevalence of severe eating disorders such as anorexia nervosa are increasing.

It is well-established that malnutrition and clinical eating disorders such as anorexia nervosa are negative for bone health. Adolescents and adults with anorexia have very low bone mass and high risk for fracture (1). Weight loss through dieting or illness causes rapid bone loss (2,3). In anorexia nervosa, multiple factors cause bone loss and/or lack of bone gain including inadequate nutrition, decreased mechanical load on bone (due to low muscle mass and low body weight), lower reproductive hormone levels, and stress-related increases in cortisol levels.

In the general population, many generally healthy individuals, consciously try to limit their food intake in order to achieve or maintain a desired body weight or size. This is referred to as dietary restraint. In our research at the University of British Columbia, we have evidence from three different groups of girls or women that dietary restraint, as well as dieting behavior, influence how much bone is gained during adolescence and increase the risk for fracture. In this article we will define eating attitudes and dieting behavior, describe how stressors act to influence bone mass, and finally summarize key studies showing negative effects of dietary restraint and dieting on bone mass.

Eating attitudes and dieting behavior

Eating attitudes are assessed in both clinical and research settings with well established questionnaires. Two of those questionnaires, which we have used in our research, include the Three Factor Eating Questionnaire (TFEQ) (4) and the children's Eating Attitudes Test (chEAT) (5). Both include questions used to define behaviors including dieting and bulimia (meaning self-induced vomiting). They also have questions designed to determine how much an individual thinks about or attempts to restrict food intake - termed "Dietary Restraint" on the TFEQ and "Oral Control" on the chEAT. Although these two terms are not identical, they are highly related and the terms will be used interchangeably for the purposes of this article.

Dietary restraint is defined as "a conscious attempt to limit food intake to regulate body weight". Individuals scoring high on the restraint scale are extremely aware of the amount and type of food they consume (6). An example question used to assess restraint is: "how likely are you to consciously eat less than you want?". Answers are scored from 1-5 ranging from not at all (1) to extremely (5). It is important to note that, although someone may score high on the restraint scale, they do not necessarily exhibit different patterns of eating behavior. In other words, many individuals who are extremely conscious and concerned about what they eat, do not actually eat less than those who happily eat what they want. In contrast, dieting behavior is an actual limiting of caloric intake resulting in weight loss. Weight loss schemes are often ineffective over the long-term and a rapid re-gain of weight is common. This loss and re-gain of weight is referred to as weight-cycling. We will discuss the effects of both eating attitudes/dietary restraint, and weight-cycling (or dieting) behavior. In our research, we have also studied individuals who score within what is considered to be the "normal" range for eating attitudes.

How could eating attitudes or dieting behavior influence bone mass?

Threats / Stresses:

• Physical — cold heat etc.
• Emotional — relationships
• Nutritional — undernutrition, fasting
• Overtraining — excess exercise

Figure 1. Process through which physical, emotional, or nutritional challenges cause increased release of CRH from the hypothalamus. These factors suppress the reproductive system and stimulate the adrenal axis. Abbreviations: ACTH = cortiotrophin; LH = lutenizing hormone. Adapted from Prior JC. Exercise associated menstrual disturbances. In: Adashi EY, Rock JA, Rosenwaks Z (eds). Reproductive Endocrinology, Surgery, and Technology. Raven, New York, 1996, pp. 1077-1091

Eating restraint and dieting behavior can be chronic stressors for many individuals. We hypothesized that bone loss occurs due to subtle decreases in reproductive hormones and increases in cortisol (a stress hormone) - both of these changes negatively effect bone mass.

In order to understand these relationships, it is important to discuss how the endocrine system responds and adapts to perceived stressors. In women, physiological adaptation to a number of stressors often results in changes in menstrual cycle function. These alterations range from subtle changes including shortened luteal phase length and anovulation in the presence of normal cycle length, to long cycles or amenorreha (lack of menstruation(. Menstrual cycle and ovulatory disturbances are often protective and necessary adaptations to increased stress (Figure 1).

The hypothalamus, which controls reproductive function and stress hormone release, receives information from several areas of the brain and functions to maintain homeostasis in response to internal and external demands. In situations of high stress, the body may adapt to prevent pregnancy for women or decrease libido and reproductive hormones in men, as a protective mechanism to conserve energy. Psychological or physiological stresses such as life changes, loss of a loved one, college stresses, inadequate energy intake, eating restraint, or rapid increases in exercise training, have all been associated with menstrual cycle disturbances.

These may be mediated by increased release of corticotrophin-releasing hormone (CRH) from the hypothalamus, which is hypothesized to suppress lutenizing hormone (LH) pulses through GnRH, possibly via the B-endorphin system, leading to suppressed reproductive capacity (7). Increases in cortisol also directly affect bone mass by increasing bone resorption. Thus, bone mass is negatively affected by both the increase in cortisol and the decrease in reproductive hormones (progesterone and estrogen) (Figure 1).

Similarly, dieting or other life stresses stimulate an increase in cortisol and a decrease in reproductive function (8-10).