Cognitive Dietary Restraint Is Associated with Stress Fractures in ...

International Journal of Sport Nutrition and Exercise Metabolism, 2005, 15, 147-159 © 2005 Human Kinetics Publishers, Inc.

Cognitive Dietary Restraint Is Associated with Stress Fractures in Women Runners Nanci S. Guest and Susan I. Barr High levels of cognitive dietary restraint (CDR) have been associated with subclinical menstrual cycle irregularities and increased cortisol levels, both of which can affect bone mineral density (BMD). Low BMD has been implicated in stress fracture risk. We assessed CDR in female runners (� 20 km/wk) with a recent stress fracture (SF) and with no stress fracture history (NSF). A sample of 79 runners (n = 38 SF, 29 � 5 y; n = 41 NSF, 29 � 6 y) completed a 3-d food record and questionnaire assessing physical activity, menstrual cycle history, and perceived stress. SF and NSF runners had similar body mass index (21.2 � 1.8 vs. 22.0 � 2.5 kg/m2), physical activity (35.7 � 13.5 vs. 33.4 � 1.34 km/wk), perceived stress, and dietary intakes. CDR, however, was higher in SF runners (11.0 � 5.4 vs. 8.4 � 4.3, P < 0.05). Subclinical menstrual cycle disturbances and increased cortisol levels that are associated with high CDR, might in turn contribute to lowered BMD and increased stress fracture risk. Key Words: exercise, athlete, female athlete triad, physical endurance, diet

The female athlete faces unique challenges, as she lives in a society that values an “ideal” body shape and competes in a sporting arena where an ideal body weight or lean appearance equates with success. The “female athlete triad” is the combination of disordered eating, menstrual irregularity, and osteoporosis/osteopenia (29), which are interrelated in etiology, pathogenesis, and consequences. Disordered eating is central to the triad, and refers to a wide spectrum of eating attitudes and behaviors used in an attempt to lose weight to achieve a low body weight or lean appearance. Although clinical eating disorders such as anorexia nervosa represent one extreme of the spectrum of disordered eating attitudes and behaviors, more subtle eating disturbances could also have implications for the bone health of women athletes. In a recent study, Cobb et al. (10) reported that disordered eating as assessed by the Eating Disorder Inventory (EDI), which measures attitudes about food and body size, was associated with low bone mineral density (BMD) in women reporting regular menstrual cycles. Bennell et al. (6) investigated the prevalence and nature of stress fractures retrospectively in a group of 53 female competitive track and field athletes. Twenty-two women had a history of stress fractures, with a total of 45 fractures reported. Significantly higher scores on the Eating Attitudes Test (EAT-40) were seen in the stress fracture group compared to the non-stress fracture group. The higher EAT-40 score (higher scores reflect higher levels of weight and eating The authors are with the University of British Columbia, Vancouver, BC V6T 1Z4. 147

Guest and Barr

148

concerns) in the stress fracture group was also supported by positive responses to other questions related to dieting and “carefulness about weight.” In a multivariate analysis, the questions regarding “carefulness about weight” were an independent predictor of a history of stress fracture (6). To date, no studies have examined a possible relationship between cognitive dietary restraint (CDR) and stress fractures in women athletes reporting regular menstrual cycles. CDR refers to the conscious efforts to limit food intake to maintain or achieve a desired body weight (40). Actual energy intake might not be decreased in women with high levels of CDR; however, they are always aware of and monitoring what and how much they eat. We have previously documented a higher prevalence of subclinical menstrual disturbances (4, 5, 24), increased urinary cortisol excretion (25), and lower BMD (26) in women with high CDR scores, despite similar body-mass index (BMI; kg/m2), than those with lower CDR scores. The constant monitoring of food intake and concerns about weight, characteristic of restrained eaters, could represent a stressor that increases cortisol, which has the potential to affect the menstrual cycle and bone health (7, 8). Based on the associations between high CDR scores and subtle menstrual irregularities, increased cortisol, and lower BMD, active women with high CDR scores might be at increased risk for stress fractures. We hypothesized that regularly menstruating female runners with a recent lower-extremity stress fracture would have higher scores on the CDR subscale of the Three-Factor Eating Questionnaire (TFEQ), compared to a sample of regularly menstruating female runners without a history of a stress fracture, but with similar activity levels, physical, and lifestyle characteristics.

Methods In this cross-sectional, descriptive study, differences between groups were analyzed. The independent grouping variable was the presence of a recent lower-extremity stress fracture, and the dependent variable was the score on the CDR subscale of the TFEQ described below. Subjects and Protocol The sampling method used in the study for both the study group and the controls was a sample of convenience. Subjects in the study group consisted of female runners diagnosed with a recent (less than 1.5 y previously) lower-extremity stress fracture, while the control group consisted of female runners without a history of a stress fracture. These groups are subsequently referred to as the stress fracture and non-stress fracture groups, respectively. Runners in the stress fracture group were recruited through the Allan McGavin Sports Medicine Centre located at the University of British Columbia, Vancouver. Those with a current stress fracture (less than 3 months previously and therefore being monitored by a physician) were recruited through posters at the centre or at the suggestion of the treating physician during a patient visit. Interested subjects contacted the primary investigator by telephone. Women with past stress fractures (more than 3 months but less than 1.5 y previously) were identified through medical chart searches and were initially contacted by letter. They contacted the primary investigator if they were interested

Dietary Restraint and Stress Fractures

149

in participating in the study. Female runners for the non-stress fracture group were recruited through announcements at running club meetings and by posters distributed on campus, to running apparel stores, and to local fitness clubs. The University of British Columbia Behavioural Research Ethics Board approved the study prior to subject recruitment, and all subjects provided written informed consent. Recruitment criteria for all runners were: age 20 to 40 y, nulliparous, menstrual cycles of normal length (21 to 35 d), stable body weight with a BMI range of 18.5 to 25 kg/m2, recreational or competitive runner (� 20 km/wk for 9 or more months per year for 1 or more years). Exclusion criteria were: cigarette use, use of medications that could affect bone (e.g., steroids), ever diagnosed or treated for an eating disorder, presently dieting (on a specific weight loss diet), diagnosed with clinical hirsutism (excessive facial hair), or consuming more than 7 drinks per wk. Use of oral contraceptives was not an exclusion criterion. The sample size was determined by power analysis calculations using a power of 0.8 and a medium effect size of 0.5 (9). The sample size calculated was to include 40 subjects in each of the 2 groups; stress fracture, and non-stress fracture. Eligible subjects met with the primary investigator and received the questionnaire and 3-d food record forms, and instructions on how to complete them. Subjects completed the questionnaire and diet record at home and were provided with a postage-paid envelope to mail their forms back to the university within 10 d. All information was confidential. Questionnaire The questionnaire was designed to take 20 to 30 min to complete. It included previously validated, standardized scales designed to assess eating behaviors, physical activity, and perceived stress. Additional questions assessed physical and lifestyle characteristics, as described below. Eating Behaviors The 51-item TFEQ (40) was used to assess 3 dimensions of human eating behavior: 1) cognitive restraint of eating (i.e., CDR), 2) disinhibition, and 3) hunger. The CDR scale (21 items) measures the intent to control food intake to achieve or maintain a desired body weight. The disinhibition scale (16 items) assesses overeating and binge eating in response to a variety of situations associated with loss of control of food intake. The hunger scale (14 items) measures perceived hunger. The CDR score of the TFEQ was the main outcome measure in this study. The first question on the TFEQ was changed from “When I smell a sizzling steak....” to “When I smell my favorite food....” to make it suitable for those individuals who do not consume meat. Responses to items on the TFEQ were scored according to instructions provided by the authors (40) and summed to obtain scores for CDR, disinhibition, and hunger. Physical Activity The Baecke Questionnaire of Habitual Physical Activity (BQHPA) consists of 3 sections: work activity, sports activity, and non-sports leisure activity (2). The questionnaire includes 16 items scored on a Likert scale, ranging from “Never” to

150

Guest and Barr

“Always” or “Very Often.” For the two most frequently reported sport activities, additional questions query the number of months per year and hours per week of participation. The work, sports, leisure, and total activity indices were calculated according to the authors’ instructions, with higher values indicating higher levels of activity by section or by total activity score. Specific questions regarding the average distance run per week and the number of years subjects had been running were also included as additional questions in the “other information” section of the questionnaire. Runners whose stress fracture occurred within the past 3 months reported their usual running activity prior to their fracture, while all others reported their current usual running activity. Perceived Stress The Perceived Stress Scale (PSS) is a measure of the degree to which situations in one’s life are appraised as stressful (11). The PSS consists of 14 items, each scored from 1 to 4, with higher scores reflecting higher perceived stress. The scale was designed to assess the degree to which respondents find their lives unpredictable, uncontrollable, and overloaded. In the present study 1 item was inadvertently excluded from the questionnaire, so a transformed score was calculated by multiplying the score by 1.077 (14/13) to account for the omission. Physical and Lifestyle Characteristics Subjects reported their present age, height, and weight, as well as the weight at which they “feel their best.” BMI was calculated from these values. Subjects were also asked whether they had ever been diagnosed with or treated for an eating disorder, to ascertain that exclusion criteria were met. Weight fluctuation was determined by the number of times that more than 5 pounds were lost in the past 2 y. Menstrual cycle information included age of menarche, whether they were currently having menstrual cycles, and if so, whether they were regular (21 to 35 d) or irregular (> 35 d). They were also asked the average length of their menstrual cycle, and if they were currently using oral contraceptives. Lifestyle information included questions regarding alcohol and caffeinated beverage use, cigarette use, as well as medication and supplement (vitamins, minerals, herbal formulas) use. Subjects were also asked if they followed a lacto-ovo vegetarian, vegan, or other special diet. Dietary Intake Subjects recorded their food and fluid intake for 3 consecutive days, including 1 weekend day. Household measures were the primary means of recording amounts; however, gram weights were used as available. In an attempt to minimize underreporting, subjects were asked to maintain their usual intakes. The importance of this to the study and to their personal feedback was emphasized. Also, the instructions mentioned foods that are typically underreported (e.g., soft drinks, snacks, alcohol, candy, condiments), and included examples of foods perceived by some as “less healthy” (e.g., cakes, cookies, hamburgers, pizza). Ambiguous entries were clarified with subjects over the telephone prior to data entry, and food intake records were analyzed using the computer program Nutritionist Five Version 1.6 (The Hearst


151

Corp., San Bruno, CA). Total energy, carbohydrate, protein, fat, fiber, calcium, vitamin D, and iron intakes were calculated and averaged over the 3 d. Statistical Analysis Statistical analyses were performed using SPSS for Windows, version 10.1 (SPSS Inc., Chicago IL). Group comparisons for all variables were made between stress fracture and non-stress fracture groups. Mean characteristics of the 2 groups were made using independent t-tests, while comparisons involving group proportions were made using 2. Logistic regression was used to determine variables that differentiated the stress fracture and non-stress fracture groups. All analyses were two-tailed and the significance level was set at P < 0.05.

Results Eighty-six women completed the study. Seven subjects (4 in the stress fracture group and 3 in the non-stress fracture group) who admitted to being diagnosed with or treated for an eating disorder (by a “yes” response in the questionnaire) were excluded from analysis. Thirty-eight female runners in the stress fracture group and 41 runners in the non-stress fracture group were used in the analysis. Descriptive, Physical Activity, Lifestyle, and Psychometric Characteristics Descriptive characteristics of the 2 groups are presented in Table 1. There were no significant differences in age, physical, and menstrual cycle characteristics (including the proportion using oral contraceptives), or BMI values between the 2 groups. Best weight and weight fluctuation were also similar between the groups, as were perceived stress scores. Total physical activity scores assessed using the BQHPA were similar for the 2 groups; the subscales for job, sport, and leisure activity were also similar when compared by section. The length of time subjects had been running and their weekly running distance were also similar between runners in the stress fracture and non-stress fracture groups. There were no differences on the TFEQ hunger and disinhibition subscales between the 2 groups. CDR scores, however, were significantly higher in the stress fracture group compared to the non-stress fracture group. There was a tendency for a greater proportion of women in the stress fracture group to have a CDR score of 13 or above (40% vs. 22%; 2 = 2.9, P = 0.09). CDR scores of the stress fracture group did not differ between those with current (n = 12) or past (n = 26) stress fractures (9.8 ± 5.5 and 11.5 ± 5.4, respectively, t = –1.1, P = 0.29). In logistic regression analysis, only 2 variables were significantly associated with the presence or absence of a recent lower-extremity stress fracture. CDR score was the first variable selected, and was positively associated with presence of a stress fracture (� � standard error = 0.145 � 0.56, Wald = 6.77, P = 0.009). Disinhibition score, which entered second, was negatively associated with stress fractures (� = –0.165 � 0.69, Wald = 5.67, P = 0.017). Overall, 67% of cases were classified correctly, and the Nagelkerke pseudo r2 was 0.182, indicating that the model accounted for 18% of the variance in presence or absence of a stress fracture.

Guest and Barr

152

Table 1 Physical and Menstrual Cycle Characteristics, Running Activity, and BQHPA, TFEQ, PSS Scores in Non-Stress Fracture and Stress Fracture Groups

Characteristics Physical and menstrual cycle characteristics Age (y) Height (cm) Weight (kg) Body-mass index (kg/m2) Best weight (kg) Times � 5 lb weight lost in last 2 y Age of menarche (y) Menstrual cycle length (d) Oral contraceptive use (%) Running activity Running distance (km/wk) Length of time running (y) Baecke Questionnaire of Habitual Physical Activity (BQHPA) scores Total activity index Job activity index Sport activity index Leisure activity index Three-Factor Eating Questionnaire (TFEQ) scores Cognitive dietary restraint Disinhibition Hunger Perceived Stress Scale (PSS) score

Non-stress fracture (n = 41) 29.1 � 5.0 165.1 � 5.3 60.2 � 7.8 22.0 � 2.5 57.9 � 6.7 1.2 � 1.2 13.0 � 1.5 28.6 � 2.7 34

Stress fracture (n = 38) 29.2 � 5.5 167.0 � 6.9 59.1 � 6.6 21.2 � 1.8 56.8 � 5.2 1.3 � 1.8 13.4 � 1.9 28.0 � 2.9 50

33.4 � 13.4 6.7 � 4.5

35.7 � 13.5 8.2 � 4.9

7.7 � 0.7 2.5 � 0.4 2.2 � 0.4 2.9 � 0.4

7.8 � 0.7 2.6 � 0.4 2.3 � 0.4 2.9 � 0.5

8.4 � 4.3 6.7 � 3.9 6.5 � 2.7 23.9 � 5.0

11.0 � 5.4* 5.1 � 3.6 6.1 � 2.6 25.1 � 4.7

Note. Data are means � standard deviation. *P < 0.05 (independent t-test)

Energy and Nutrient Intakes Subjects’ daily energy and nutrient intakes are presented in Table 2. Nutrient intake values for vitamins and minerals include intakes from both diet and supplements. Intakes of energy, protein, carbohydrate, fat, fiber, and iron were similar between the 2 groups. Women in the stress fracture group, however, had significantly higher total daily calcium and vitamin D intakes compared to the women in the non-stress fracture group. Consumption of caffeinated beverages (daily) and alcoholic beverages (weekly) were similar between the 2 groups (Table 2). Approximately onethird of the subjects were using a vitamin, mineral, or herbal supplement, with no differences between the 2 groups. A vegetarian diet was followed by approximately one-quarter of the subjects, with no differences between the 2 groups.


153

Table 2 Daily Energy and Nutrient Intakes (from 3-D Diet Records of Food and Supplements), Vegetarianism, and Supplement Use in Non-Stress Fracture and Stress Fracture Groups

Characteristics

Non-stress fracture (n = 41)

Stress fracture (n = 38)

Energy (kcal) Protein (g) % energy Carbohydrate (g) % energy Fat (g) % energy Fiber (g) Total calcium (mg) Total vitamin D (IU) Total iron (mg) Caffeinated beverage use (cups/d) Alcoholic beverage use (drinks/wk) Vitamin, mineral, or herbal supplement use (%) Vegetariana (%)

1948 � 317 80 � 18 16.5 � 3.1 278 � 74 56.7 � 8.8 57 � 15 26.6 � 6.0 26 � 10 1024 � 361 105 � 74 18 � 7 1.6 � 1.3 2.2 � 2.4 37 24

1920 � 375 84 � 23 17.6 � 4.3 270 � 73 55.9 � 7.5 54 � 14 25.4 � 5.4 23 � 8 1290 � 524* 214 � 212* 21 � 15 1.5 � 0.9 1.6 � 2.6 34 24

Note. Data are expressed as mean � standard deviation. *P < 0.05 (independent t-test). a Vegetarian refers to those who exclude meat and poultry. Two vegetarians consumed fish occasionally.

Discussion This is the first study to report an association between CDR and stress fractures in women runners. We hypothesized that a sample of regularly menstruating runners with a recent lower-extremity stress fracture would have higher CDR scores, compared to a sample of regularly menstruating runners without a history of a stress fracture, but with similar activity level, and physical and lifestyle characteristics. In our study, runners in the stress fracture and non-stress fracture groups had similar BMI, physical activity, perceived stress, and energy and macronutrient intakes. Our main finding was significantly higher CDR, as assessed by the TFEQ, in runners with a recent stress fracture compared to runners without a history of a stress fracture. Stress fractures are common overuse injuries in female runners, and it is well documented that stress fractures are associated with eating disorders and amenorrhea/oligomenorrhea (29, 41), as well as osteopenia or low BMD (18, 28). These interrelated conditions have been referred to as the “female athlete triad” (29), usually when the symptoms of each component reach a clinically recognizable pathological state. Therefore, less obvious subclinical levels of the triad components might be overlooked because of their non-symptomatic nature. Less severe forms

154

Guest and Barr

of disordered eating and menstrual irregularities, however, could in fact predispose female athletes to an increased risk for stress fracture or suboptimal long-term bone health, including premature osteoporosis. One mechanism that might underlie this association is related to the possibility that daily decisions about what and how much to eat represent a chronic, low-level stressor in women with high levels of CDR (Figure 1). Stress, whether physical or psychological, stimulates the release of corticotropin releasing-hormone (CRH) from the hypothalamus, which ultimately leads to cortisol release from the adrenal gland (37). Higher levels of cortisol have been detected in women with high versus low CDR (1, 25) and appear to be associated with menstrual disturbances and lower BMD or bone mineral content (26, 43). Specifically, high cortisol concentrations are associated with reproductive disturbances resulting from the inhibitory effect of CRH on hypothalamic and hence pituitary secretions required for normal menstrual function (3, 7). Reduced exposure to the female reproductive hormones estradiol and progesterone, which occurs in oligo/amenorrhea, is associated with lower spinal BMD (12, 13, 22, 27, 41) and with lower BMD of whole body and appendicular sites (27, 31, 36) in female athletes. More subtle reproductive disturbances, however, such as anovulatory cycles and shortened luteal phases, have also been associated with spinal trabecular bone loss (34). These subclinical disturbances occur more frequently in women with high levels of CDR (4, 5, 19, 24, 39). In addition to its associations with reproductive disturbances, cortisol is known to have direct effects on bone (8). It is well documented that excessive glucocorticoid levels, both from endogenous (14) as well as therapeutic exogenous sources (44), are implicated in bone loss. Preliminary evidence has suggested that high CDR scores are associated with lower BMD or bone mineral content (26, 43). Low BMD might contribute to the development of stress fractures by

Figure 1—Proposed pathways among cognitive dietary restraint, the menstrual cycle, bone mineral density and stress fracture risk. Solid lines represent associations suggested by previous studies; dashed line represents association suggested by current study.


155

reducing bone strength and allowing the microdamage from repetitive loading to accumulate; studies have reported an association between low BMD and stress fracture risk (18, 28). This proposed mechanism is unlikely to be specific to high levels of CDR as a marker for disturbed eating attitudes. The data of Cobb et al. (10) indicate that high scores on 3 subscales of the EDI were also associated with lower BMD in eumenorrheic women runners of normal weight, while Bennell et al. (6) found that women with a history of stress fractures had higher scores on the EAT-40 and questions on “carefulness about weight.” Moreover, the proposed mechanism could apply to any type of chronic stress, rather than being restricted to stress associated with concerns about body weight and shape. Among young women athletes, however, these concerns are so prevalent as to be almost normative, and might not even be perceived as a source of stress. In this regard, it is noteworthy that the athletes in the stress fracture group of our study did not have higher scores on the Perceived Stress Scale. Although runners with stress fractures could be more likely to have eating attitudes that could increase stress, intakes of energy, macronutrients, fiber, and iron were similar between the stress fracture and non-stress fracture groups. Women in the stress fracture group, however, had significantly higher total daily calcium and vitamin D intakes compared to women in the non-stress fracture group. We suggest that the reason for the group differences in vitamin D and calcium intake was the treating physician’s recommendations to increase these bone-specific nutrients at the time of stress fracture diagnosis. Most of the runners reported their dietary intake weeks or months after their diagnosis, when dietary changes (i.e., increased dietary or supplemental vitamin D and calcium) would have presumably been implemented. Runners diagnosed with a stress fracture would have had greater awareness of the importance of these nutrients to bone health, compared to the non-stress fracture group. Underreporting is pervasive in dietary intake studies (42) and undoubtedly occurred to some extent among our subjects. It is unlikely, however, that differences in energy availability between the stress fracture and non-stress fracture groups confounded our results, because energy intakes, body weights, and activity levels were similar. Furthermore, the extent of underreporting did not appear excessive: mean reported intake for the entire group was 1935 kcal/d, which was reasonably close to the mean estimated energy expenditure of 2127 kcal/d (calculated based on individual height, weight, and age, and an average running distance of 5 km/d in addition to activities of daily living) (15). Limitations of our study include the absence of measured data on BMD, cortisol levels, and subclinical menstrual disturbances in our subjects. Therefore, it is possible that CDR scores were increased as a consequence of experiencing stress fractures rather than being associated with their occurrence. Although this possibility cannot be excluded, we believe that it is unlikely. First, a mechanism for this association exists: both our group and others have documented associations among CDR and subclinical menstrual disturbances, cortisol, and BMD, as previously described. Second, an opportunity to examine this issue was provided by the fact that our stress fracture group included women currently recovering from a stress fracture, as well as recovered individuals who had resumed running. One might speculate that CDR scores in women with a current stress fracture would

Guest and Barr

156

increase to compensate for decreased activity during recovery (although in most cases these women were remaining active by participating in non-weight-bearing activities). A comparison of CDR scores between runners whose stress fractures had occurred within the past 3 months to those who had fully recovered revealed that CDR scores were not different, however. Other potential limitations of the study are the absence of full menstrual history data and data on the reason for oral contraceptive use. We have previously shown that high levels of CDR are associated with subclinical menstrual disturbances and irregular cycles (4, 5, 24); thus it is possible that women in the stress fracture group (with higher CDR scores) might have been more likely to have experienced episodes of amenorrhea or oligomenorrhea in the past. Similarly, although the prevalence of oral contraceptive use did not differ between groups, it is conceivable that women in the stress fracture group could have been more likely to use them to treat irregular or absent cycles, rather than for fertility control. Neither of these possibilities would be inconsistent with our hypothesis. Thus, while these additional data might have added to our results, their absence does not invalidate our findings. Further, with regard to oral contraceptive use, it is not clear how a difference in the reason for use between groups would have influenced BMD and risk of stress fractures: studies have yielded mixed results as to whether oral contraceptive use is associated with higher (17, 20, 30), lower (32, 33), or no difference in BMD (16, 21, 23, 35, 38) in premenopausal women. In conclusion, we found higher levels of CDR in women runners with a recent stress fracture compared to those without a history of a stress fracture. This observation is a concern for women runners, as stress fractures are presently a common injury in this population. High levels of CDR or other types of disordered eating attitudes (and the corresponding likelihood of subclinical menstrual disturbances and increased cortisol) might be additional risk factors for stress fractures that are overlooked by both the athlete herself as well as health care providers, because of their non-symptomatic nature. Acknowledgments We thank the women who participated in our study for their time and interest. We also appreciate the assistance of Jack E. Taunton, M.D., in recruiting subjects with stress fractures and thank both Dr. Taunton and Judy A. McLean, Ph.D., for their contributions during the course of the study.

References 1. Anderson, D.A., J.R. Shapiro, J.D. Lundgren, L.E. Spataro, and C.A. Frye. Self reported dietary restraint is associated with elevated levels of salivary cortisol. Appetite. 38:1317, 2001. 2. Baecke, J.A.H., J. Burema, and J.E.R. Frijters. A short questionnaire for the measurement of habitual activity in epidemiological studies. Am. J. Clin. Nutr. 36:936-942, 1982. 3. Barbarino, A., L. De Marinis, G. Folli, A. Tofani, S. Della Casa, C. D’Amico, A. Mancini, S.M. Corsello, P. Sambo, and A. Barini. Corticotropin-releasing hormone inhibition of gonadotropin secretion during the menstrual cycle. Metabolism. 38: 504-506, 1989.


157

4. Barr, S.I., J.C. Prior, and Y.M. Vigna. Restrained eating and ovulatory disturbances: possible implications for bone health. Am. J. Clin. Nutr. 59:92-97, 1994. 5. Barr, S.I., K. Janelle, and J.C. Prior. Vegetarian vs. non-vegetarian diets, dietary restraint and subclinical ovulatory disturbances: prospective six-month study. Am. J. Clin. Nutr. 60:887-894, 1994. 6. Bennell, K.L., S.A. Malcolm, S.A. Thomas, P.R. Ebeling, P.R. McCrory, J.D. Wark, and P.D. Brukner. Risk factors for stress fractures in female track and field athletes: a retrospective analysis. Clin. J. Sports Med. 5:229-235, 1995. 7. Berga, S.L. Stress and ovarian function. Am. J. Sports Med. 24(6 suppl):S36-S37, 1996. 8. Canalis, E. Mechanisms of glucocorticoid action in bone: implications to glucocorticoid-induced osteoporosis. J. Clin. Endocrinol. Metab. 81:3441-3447, 1996. 9. Cheney, C.L., and C.J. Boushey. Estimating sample size. In: Research: Successful Approaches. E. Monsen (ed.) Chicago, IL: American Dietetic Association, pp. 337346, 1992. 10. Cobb, K.L., L.K. Bachrach, G. Greendale, R. Marcus, R.M. Neer, J. Nieves, M.F. Sowers, B.W. Brown, G. Gopalakrishnan, C. Luetters, H.K. Tanner, B. Ward, and J.L. Kelsey. Disordered eating, menstrual irregularity, and bone mineral density in female runners. Med. Sci. Sports Exerc. 35:711-719, 2003. 11. Cohen, S., T. Kamarck, and R. Mermelstein. A global measure of perceived stress. J. Health Soc. Behav. 24:385-396, 1983. 12. Drinkwater, B.L., B. Bruemner, and C.H. Chesnut 3rd. Menstrual history as a determinant of current bone density in young athletes. JAMA. 263:545-548, 1990. 13. Drinkwater, B.L., K. Nilson, C.H. Chesnut, W.J. Bremner, S. Shainholtz, and M.B. Southworth. Bone mineral content of amenorrheic and eumenorrheic athletes. N. Engl. J. Med. 311:277-281, 1984. 14. Greendale, G.A., J.B. Unger, J.W. Rowe, and T.E. Seeman. The relation between cortisol excretion and fractures in healthy older people: results from the MacArthur studies. J. Am. Geriatr. Soc. 47:799-803, 1999. 15. Institute of Medicine, Food and Nutrition Board. Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein and Amino Acids. National Academy Press: Washington, DC: 2002. 16. Keay, N., I. Fogelman, and G. Blake. Bone mineral density in professional female dancers. Br J Sports Med 31:143-147, 1997. 17. Laitinen K., M. Valimaki, and P. Keto. Bone mineral density measured by dual energy xray absorptiometry in healthy Finnish women. Calcif. Tissue Int. 48:224-231, 1991. 18. Lauder, T.D., S. Dixit, L.E. Pezzin, M.V. Williams, C.S. Campbell, and G.D. Davis. The relation between stress fractures and bone mineral density: evidence from active-duty army women. Arch. Phys. Med. Rehabil. 81:73-79, 2000. 19. Lebenstedt, M., P. Platte, and K.M. Pirke. Reduced resting metabolic rate in athletes with menstrual disorders. Med. Sci. Sports Exerc. 31:1250-1256, 1999. 20. Lindsay, R., J.F. Tohme, and B. Kanders. The effect of oral contraceptive use on vertebral bone mass in pre- and post-menopausal women. Contraception. 34:333-340, 1986. 21. Lloyd T., J.R. Buchanan, G.R. Ursino, C. Myers, G. Woodward, and D.R. Halbert. Long-term oral contraceptive use does not affect trabecular bone density. Am. J. Obstet. Gynecol. 160:402-404, 1989. 22. Marcus, R., C. Cann, P. Madvig, J. Minkoff, M. Goddard, M. Bayer, H. Martin, L. Guadiani, W. Haskell, and H. Genant. Menstrual function and bone mass in elite women distance runners. Endocrine and metabolic features. Ann. Intern. Med. 102:158-163, 1985.

158

Guest and Barr

23. Mazess R.B., and H.S. Barden. Bone density in premenopausal women: effects of age, dietary intake, physical activity, smoking, and birth-control pills. Am. J. Clin. Nutr. 53: 132-142, 1991. 24. McLean, J.A., and S.I. Barr. Cognitive dietary restraint is associated with eating behaviors, lifestyle practices, personality characteristics and menstrual irregularity in college women. Appetite. 40:185-192, 2003. 25. McLean, J.A., S.I. Barr, and J.C. Prior. Cognitive dietary restraint is associated with higher urinary cortisol excretion in healthy premenopausal women. Am. J. Clin. Nutr. 73:7-12, 2001. 26. McLean, J.A., S.I. Barr, and J.C. Prior. Dietary restraint, exercise, and bone density in young women: are they related? Med. Sci. Sports Exerc. 33:292–296, 2001. 27. Micklesfield, L.K., E.V. Lambert, A.B. Fataar, T.D. Noakes, and K.H. Myburgh. Bone mineral density in mature, premenopausal ultramarathon runners. Med. Sci. Sport Exerc. 27:688-696, 1995. 28. Myburgh, K.H., H.J. Hutchins, A.B. Fataar, S.F. Hough, and T.D. Noakes. Low bone density is an etiologic factor for stress fractures in athletes. Ann. Intern. Med. 113:754759, 1990. 29. Otis, C.L., B. Drinkwater, M. Johnson, A. Loucks, and J. Wilmore. American College of Sports Medicine position stand. The Female Athlete Triad. Med. Sci. Sport Exerc. 29:i-ix, 1997. 30. Pasco, J.A., M.A. Kotowicz, M.J. Henry, S. Panahi, E. Seeman, and G.C. Nicholson. Oral contraceptives and bone mineral density: a population-based study. Am. J. Obstet. Gynecol. 182:265-269, 2000. 31. Pettersson, U., B. Stalnacke, G. Ahlenius, K. Henriksson-Larsen, and R. Lorentzon. Low bone mass density at multiple skeletal sites, including the appendicular skeleton in amenorrheic runners. Calcif. Tissue Int. 64:117-125, 1999. 32. Polatti F., F. Perotti, N. Filippa, D. Gallina, and R.E. Nappi. Bone mass and long-term monophasic oral contraceptive treatment in young women. Contraception. 51:221-224, 1995. 33. Prior, J.C., S.A. Kirkland, L. Joseph, N. Kreiger, T.M. Murray, D.A. Hanley, J.D. Adachi, Y.M. Vigna, C. Berger, L. Blondeau, S.A. Jackson, and A. Tenenhouse. Oral contraceptive use and bone mineral density in premenopausal women: cross-sectional, population-based data from the Canadian Multicentre Osteoporosis Study. Can. Med. Assoc. J. 165:1023-1029, 2001. 34. Prior, J.C., Y.M. Vigna, M. Schechter, and A. Burgess. Spinal bone loss and ovulatory disturbances. N. Engl. J. Med. 323:1221-1227, 1990. 35. Reed, S.D., D. Scholes, A.Z. LaCroix, L.E. Ichikawa, W.E. Barlow, and S.M. Ott. Longitudinal changes in bone density in relation to oral contraceptive use. Contraception. 68:177-182, 2003. 36. Rencken, M.L., C.H. Chesnut, and B.L. Drinkwater. Bone density at multiple sites in amenorrheic athletes. JAMA. 276:238-240, 1996. 37. Rivier, C., J. Rivier, and W. Vale. Stress-induced inhibition of reproductive functions: role of endogenous corticotropin-releasing factor. Science. 231:607-609, 1986. 38. Rodin A., M. Chapman, and I. Fogelman. Bone density in users of combined oral contraception. Br. J. Fam. Plann. 16:125-129, 1991. 39. Schweiger, U., R. J. Tuschl, P. Platte, A. Broocks, R.G. Laessle, and K.M. Pirke. Everyday eating behaviour and menstrual function in young women. Fertil. Steril. 57: 771-775, 1992.


159

40. Stunkard, A.J., and S. Messick. The three-factor eating questionnaire to measure dietary restraint, disinhibition and hunger. J. Psychosom. Res. 29:71-83, 1985. 41. Tomten, S.E., J.A. Falch, K.I. Birkeland, P. Hemmersbach, and A.T. Høstmark. Bone mineral density and menstrual irregularities. A comparative study on trabecular bone structures in runners with alleged normal eating behaviour. Int. J. Sports Med. 19:92-97, 1998. 42. Trabulsi, J., and D.A. Schoeller. Evaluation of dietary assessment instruments against doubly labeled water, a biomarker of habitual energy intake. Am. J. Physiol. Endocrinol. Metab. 281:E891-E899, 2001. 43. Van Loan, M.D., and N.L. Keim. Influence of cognitive eating restraint on total-body measurements of bone mineral density and bone mineral content in premenopausal women aged 18-45y: a cross-sectional study. Am. J. Clin. Nutr. 72:837-843, 2000. 44. Ziegler, R., and C. Kasperk. Glucocorticoid-induced osteoporosis: prevention and treatment. Steroids. 63(5-6):344-348, 1998.