|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Diet and Breast Cancer: Evidence That Extremes in Diet Are Associated With Poor Survival
From the Department of Medicine, Department of Surgery, Division of Clinical Epidemiology, Samuel Lunenfeld Research Institute Mount Sinai Hospital, Toronto-Sunnybrook Regional Cancer Centre, St. Michael’s Hospital, University of Toronto, Toronto, Canada. Address reprint requests to Pamela J. Goodwin, MD, Mount Sinai Hospital, 1284-600 University Ave, Toronto, Ontario M5G 1X4, Canada; email: pgoodwin{at}mtsinai.on.ca.
Purpose: Diet has been postulated to influence breast cancer prognosis; however, existing evidence is weak and inconsistent. Previous studies have sought evidence of a linear relationship between diet and breast cancer outcomes. Because of a U-shaped association of body mass index (BMI) with survival in breast cancer, we hypothesized that a nonlinear association also existed for dietary variables. Patients and Methods: Four hundred seventy-seven women with surgically resected T1 to T3, N0/1, M0 breast cancer completed the Block Food Frequency Questionnaire 9.3 ± 4.6 weeks (mean ± standard deviation) after diagnosis, reporting intake over the preceding 12 months. Data on tumor-related factors, treatment, and outcomes were obtained prospectively from medical records. A series of Cox models was performed, modeling the association of dietary factors with breast cancer survival linearly and quadratically, adjusting for total energy intake, tumor- and treatment-related variables, and BMI. Results: Significant nonlinear survival associations were found for protein, oleic acid, cholesterol, polyunsaturated-saturated fat ratio, and for percentage of calories from fat and percentage of calories from carbohydrates in multivariate models. The shape of the survival associations varied across nutrients. Hazard ratios for highest risk quintiles ranged from 2.1 to 6.5. For total fat, adjustment for BMI reduced the multivariate P value obtained from nonlinear Cox models from .05 to .10. No significant linear associations were identified. Conclusion: The association of key dietary variables with breast cancer survival may be U-shaped rather than linear. Our data suggest that midrange intake of most major energy sources is associated with the most favorable outcomes, and extremes are associated with less favorable outcomes.
IT HAS long been postulated that diet, particularly fat intake, contributes to the development and progression of breast cancer. However, the evidence relating diet to breast cancer risk is weak and inconsistent,1,2 and randomized trials are ongoing.3–5 More than a dozen observational studies6–18 have examined the association of diet with breast cancer recurrence and survival. Results have been largely negative. There is weak and inconsistent evidence that greater intake of fat, or certain types of fat, may be associated with an increased risk of recurrence or death; however, most associations have been nonsignificant, particularly after adjustment for total energy intake.6–18 Some of the factors contributing to these largely negative findings may include small sample size, measurement of diet years before breast cancer diagnosis, failure to adjust for total energy intake,19 and the sizable measurement error inherent in measuring diet or the true absence of a prognostic effect of diet.20 One of the consistent characteristics of the published studies is that evidence for linear associations of diet with breast cancer outcomes was sought. That is, data were analyzed to see if there were significant monotonic increments in risk with increasing or decreasing intake of dietary variables. Although this is one plausible biologic model, it is not the only model. For at least one other nutrition-related variable, body mass index (BMI), the association with breast cancer outcome is U-shaped (quadratic), rather than linear, with the best outcomes being seen in individuals with an intermediate BMI and individuals with either higher or lower BMIs having worse outcomes.21,22 Put in different terms, midrange BMI (20–25 kg/m2) at the time of breast cancer diagnosis has been associated with optimal survival, and extremes of BMI have been associated with less favorable survival. A similar pattern of BMI with general health outcomes is seen in the general population.23 It is possible that this may relate, at least in part, to confounding by weight loss associated with chronic disease and the dying process, although in our earlier report of BMI and breast cancer outcomes, most recurrences occurred years after diagnosis, suggesting this was not a major factor. We believe that a similar association may also be present for diet and breast cancer outcomes. That is, there may be optimal ranges of dietary intake that are associated with the best outcomes, and extremes may be associated with worse outcomes. This is consistent with observations that certain ranges of physiologic factors (eg, hemoglobin, blood electrolytes, body temperature) are associated with optimal health. Higher or lower levels are associated with reduced health. This may be particularly true in the case of diet, where high intake of one component of the diet may be associated with lower intake of other components of the diet, so that opposite extremes in intake of multiple dietary components often coexist. On reviewing published studies that have examined associations of diet with breast cancer outcomes and provided hazard ratio (HR) estimates by quantiles, there is some evidence for U-shaped prognostic effects. An early prospective cohort study by Rohan et al11 reported no significant linear association of fat intake (total fat, saturated fat, monounsaturated fat, polyunsaturated fat) with outcomes and concluded that dietary fat did not influence outcomes. The HRs presented in the report show a U-shaped pattern for these fat-related factors, the best outcomes being associated with intermediate levels of intake. A similar pattern is present for protein. A recent report (arising from the Nurses’ Health Study) by Holmes et al18 also reports no significant linear association of fat with breast cancer outcomes. However, inspection of the data reveals some evidence for a U-shaped association of animal and saturated fat and of several specific fatty acids, including linoleic and oleic acids with breast cancer outcomes. A similar pattern is observed in the data presented by Ingram et al.14 The statistical significance of these U-shaped associations can be ascertained only through reanalysis of the data using quadratic modeling. In this report, we examine the association of major dietary sources of energy and of key fat-related variables with breast cancer survival, modeling these associations quadratically as our primary analysis. Second, to facilitate comparison of our results to previous publications, we present results of analyses of linear associations. In our analyses, we seek to ascertain whether intermediate levels of dietary intake are associated with optimal breast cancer survival.
Population Assembly A consecutive cohort of women who underwent treatment for operable breast cancer at three University of Toronto hospitals (Mount Sinai, Women’s College, and St Michael’s) was assembled prospectively from July 1989 through June 1996. Details of study methods have been published elsewhere.21 They are summarized briefly here. Women met the following criteria: (1) age younger than 75 years, and (2) complete resection (lumpectomy with margins clear of invasive cancer or mastectomy) and axillary node dissection for T1 to T3, N0/1, M0 breast cancer. Exclusion criteria included the following: (1) prior malignancy (except nonmelanoma carcinoma of skin or carcinoma-in-situ of cervix), (2) serious coexisting medical condition including known diabetes, Type I or II, (3) medications that could influence diet or lipids, or (4) inability to speak English. Only premenopausal women were recruited for the first 3 years of the study; thereafter, both pre- and postmenopausal women were recruited. All participants provided written informed consent in accordance with the Human Subjects Committee of the University of Toronto.
Measurement Pathologic characteristics of the tumors were abstracted from pathology reports. Hormone receptors were measured using protein binding or immunohistochemical assays according to the standard practice at each institution.
Follow-Up
Statistical Analysis Survival analyses were performed using the Cox proportional hazards model. All models were adjusted for total caloric intake by including total calories in the model.25 Raw hazard ratio estimates were obtained from the Cox model by categorizing the dietary variables into quintiles. In calculating the HRs, the quintile with the lowest risk was assigned a value of 1, and the HR for other quintiles was calculated relative to that quintile. The prognostic effect of each diet variable x was first modeled as a continuous linear function ß1x. A likelihood ratio P value for linear trend was calculated after adjustment for age at diagnosis, adjuvant chemotherapy, adjuvant tamoxifen, tumor stage, nodal stage (results were similar when number of involved axillary nodes was used), and BMI (modeled quadratically), in addition to total calories. The same multivariate model was then fitted, but with the diet variable modeled as a continuous quadratic function consisting of a linear term ß1x plus a quadratic term ß2x2. Likelihood ratio P values were calculated for the quadratic function ß1x + ß2x2 as a whole and for the contribution of the quadratic term ß2x2 over and above the linear term. In every case that the overall P value for the quadratic function was significant, the P value for the quadratic term alone was also significant. As a result, only the former are reported here. Models adjusted only for total calories were also fitted, but results are not reported because all variables significant in these models were also significant in the multivariate models. Model-based HRs were calculated for the diet variables in the multivariate models by keeping the other variables fixed at an arbitrary point and expressing the hazard relative to the lowest hazard. If significance is mentioned without the provision of specific P values, it is at the 5% level. No adjustments were made for multiple testing.
Characteristics of the Study Population Mean ± SD age was 50.4 ± 9.8 years. Most women were premenopausal (57.7%), and approximately half had a BMI greater than 25 kg/m2, the upper limit considered optimal for good health. Most underwent lumpectomy (76.7%), and most received adjuvant chemotherapy (28.3%), hormone therapy (29.6%), or both (9.6%). Most tumors were stage T1 (55.6%), node-negative (69.4%), and estrogen receptor (ER)–or progesterone receptor (PgR)–positive (62.5% and 56.6%, respectively; Table 1  ). Median follow-up of the survivors was 6.1 years (range, 0.03–11.3 years); 52 (11%) of women had died.
Diet at Breast Cancer Diagnosis Mean intake and range of intake of key dietary variables is shown in Table 2 . Mean daily caloric intake was 1,847 kcal, distributed as 38.7% fat, 14.5% protein, 44.6% carbohydrate, and 2.9% alcohol. Strong correlations existed between total calories and fat (total fat, saturated fat, linoleic acid, oleic acid), carbohydrate, protein, and cholesterol (r = 0.70 to 0.88), as well as between percentage of calories from carbohydrates and percentage of calories from fat (inverse, r = -0.81). Correlations of total caloric intake with BMI or alcohol intake were weak (r <0.20). Alcohol accounted for 2.9% of caloric intake, intake averaging just less than one beverage every other day.
Prognostic Effects of Diet Prognostic effects of diet are also shown in Table 2  . The observed energy-adjusted HRs of death are shown first. These HRs are shown for descriptive purposes; no tests of significance were performed on them, and it is not our intention that hazard ratios for one quintile be directly compared with those of another quintile. What we are attempting to show here is the observed pattern of HRs across the range of intake of each variable. These HRs are not adjusted for treatment or other prognostic factors and, as a result, should not be directly compared with the adjusted HRs shown in the right hand side of Table 2 or portrayed in Figure 1. Looking at observed energy-adjusted quintile HRs, with one exception (carbohydrates, grams/day [g/d]) the lowest-risk quintile was Q3 or Q4. Quintiles varied in range, reflecting uneven distribution of intake. When categories were redefined to have more equal-sized ranges of nutrient intake, the same pattern persisted, although HR estimates were larger (data not shown). Inspection of these descriptive data suggests that a U-shaped association of these dietary factors with survival may be present. This was explored in the quadratic modeling below.
In Table 2 , we also provide results of two different approaches to statistical analysis of prognostic effects. First, prognostic effects are modeled linearly, and quintile HRs and P values (linear trend) are calculated after adjustment for total calories, BMI, and tumor- and treatment-related variables. This linear analysis is similar to analyses reported in prior studies. Next, prognostic associations are modeled quadratically, allowing a U-shaped association between intake and breast cancer outcomes after adjustment for total calories, BMI, and tumor- and treatment-related variables. The P values in these quadratic analyses reflect the prognostic significance of the overall quadratic function (linear plus quadratic terms). In every case where the overall P value for the quadratic function is significant, the contribution of the quadratic term itself was also significant. HRs and curves for prognostic associations generated from these multivariate quadratic Cox models are shown in Table 2 and Figure 1, respectively. For both linear and quadratic multivariate models, HRs (adjusted for energy and other prognostic factors) are smaller than the HRs that were adjusted for energy only. That is, prognostic effects of diet were attenuated after consideration of effects of other prognostic factors.
The shape of the HR curves shown in Figure 1
It can be seen that there was little evidence of significant linear associations with breast cancer outcomes. There were borderline results for protein (g/d; P = .07) and for percentage of calories from protein (P = .06), higher intake being associated with lower risk. These results may be explained by the shape of the HR curves in Figure 1C When quadratic relationships were modeled, strikingly different results were obtained. After adjustment for age, total calories, BMI, tumor and nodal stage, adjuvant hormone therapy, and adjuvant chemotherapy, significant associations were seen for protein (g/d; P = .01), oleic acid (g/d; P = .03), polyunsaturated-to-saturated fat ratio (P = .004), cholesterol (mg/d; P = .02), percentage of calories from fat (P = .03), and percentage of calories from carbohydrate (P = .002). A borderline effect was seen for total fat (g/d; P = .10) and for percentage of calories from protein (P = .10). Our decision to include BMI (modeled quadratically) in our multivariate analyses reflects previous work by ourselves21 and others.22 Specifically, we identified a strong prognostic effect of BMI in this cohort (P < .001). Inclusion of BMI in these multivariate quadratic models led to slight attenuation of P values; the significance of the diet variables remained unchanged with the exception of total fat intake (g/d; multivariate P increased from .05 when BMI was not included to .10 when BMI was included). There was no evidence of a significant interaction between BMI and total fat (P = .54). Similar survival analyses were conducted for intake of fish or chicken, any vegetable, deep yellow or dark green vegetables, any fruit or juice, fiber, vitamin A, vitamin C, and beta-carotene. No significant effects were seen (data not shown).
As predicted, results of analyses that assumed a linear relationship between diet and breast cancer survival yielded largely negative results. With the exception of a suggestion of longer survival in women whose protein intake was high, there was little evidence for an important linear prognostic effect. Had this been the only analysis we performed, our results would have concurred with previously published studies, and we would have concluded that diet around the time of breast cancer diagnosis was not associated with survival. When we inspected the pattern of HRs across quintiles, we found qualitative evidence for a U-shaped association of diet with breast cancer survival. Furthermore, when we performed quadratic modeling, we found evidence of significant survival associations for many of the dietary factors we studied. Inclusion of BMI (a factor associated with survival in this cohort) had little effect on the prognostic associations we identified, apart from dietary fat, where the significance of the quadratic association of fat intake (g/d) was attenuated when BMI was included in the model. These observations suggest that the association of diet with breast cancer survival may be more complex than was previously thought. A simple linear association was not present for most variables, whereas quadratic or U-shaped associations were identified for some variables, with midrange intake of most factors being associated with optimal survival. The shape of these associations, as generated from our Cox models, varied across dietary variables. Because these precise shapes were not hypothesized a priori, we recommend that similar analyses be conducted in other datasets to replicate our observations. Nonetheless, our results suggest that a balanced diet, with intermediate levels of intake of the major sources of energy and types of fat, may be associated with optimal breast cancer survival and that dietary extremes may be associated with worse outcomes.
Specifically, for dietary fat intake, there was evidence of a significant quadratic association when fat intake was expressed as percentage of calories. Similar patterns were seen for oleic acid, polyunsaturated-to-saturated fat ratio, and cholesterol. The prognostic significance of fat intake, expressed in grams per day, was present only when BMI was not included in the model. As can be seen in Figure 1A For protein, quintile four (15.1%-16.6% calories, 70.6–84.9 g/d) was associated with the lowest risk of death. In addition to the presence of a significant quadratic association, there was also evidence of a linear association when protein intake was expressed as grams per day that was of borderline statistical significance. This reflects a linear portion of the HR curve; our data suggest that increasing intake beyond 15% of calories or 80 g/d is associated with a fairly constant risk of death. Improved breast cancer outcomes with high protein intake has been reported by two other groups using linear models.16,18
For carbohydrates, there was a near monotonic decrease in risk with increasing intake when intake was measured as grams per day; however, there was no evidence of a significant linear trend in either univariate or multivariate analysis. In contrast, for percentage of calories from carbohydrates, the middle category of intake (42.5%-46.5%) was associated with the lowest risk of death, and a highly significant quadratic association was identified, even after multivariate adjustments. The somewhat different shapes of the HR curves (Fig 1B A modest level of alcohol intake does not seem to adversely affect survival. Our observations are in keeping with nonsignificant patterns identified by a number of previous investigators,11,12,15,18 but they are not in keeping with one report16 in which intake of beer was associated with a significantly increased risk of recurrence and death. Total energy intake was not associated with survival in any of our analyses, despite our earlier observation that BMI, modeled quadratically, predicted both recurrence and death.21 This reflects the weak correlation between BMI and total calories (Spearman r = 0.10) and highlights the importance of dietary composition rather than total energy intake in breast cancer prognosis. Our decision to use quadratic modeling arose from our observation of a U-shaped association of BMI, with breast cancer outcomes.21 We hypothesized that a similar association might be present for dietary variables. We have found evidence for such an association for many of the variables we studied. Our results suggest that women with newly diagnosed breast cancer who consume a balanced diet, avoiding extremes in intake, may have the best outcomes. Furthermore, they suggest that modest alcohol intake is not associated with adverse outcomes. We have not directly examined the effects of altering diet after breast cancer is diagnosed. Our findings should be viewed as hypothesis generating. They require replication in other datasets and do not provide sufficient evidence to recommend dietary changes in women with breast cancer. Although this was a prospective study, there are several potential limitations that may have influenced our results. Median follow-up was relatively short (just over 6 years) and the number of deaths was low (52 deaths); this may have led to reduced power to detect some prognostic associations. Long-term follow-up is underway to allow an examination of long-term effects of diet. Furthermore, the diet questionnaire required women to recall food intake during the previous year. This may have introduced a recall bias or increased measurement error, despite evidence of validity of the questionnaire when used in this fashion.24 Because of these limitations, we recommend that other investigators attempt to replicate our results, seeking evidence of quadratic or nonlinear associations of diet with breast cancer outcomes and exploring the shape of the prognostic associations. If our results are confirmed, identification of optimal ranges of intake, perhaps through pooling of data across studies, would be useful. Ideally, this would also include an exploration of joint effects of dietary variables, something that we did not undertake in our analysis involving more than 477 women. In particular, such analyses could attempt to explore whether some of the prognostic effects we have identified reflect reciprocal effects of a second nutrient when intake of the primary nutrient is changed (eg, increase in carbohydrate intake when fat intake is reduced). This might, in future, lead to a better understanding of what an optimal diet may be for women diagnosed with breast cancer. Similar analyses of the association of diet with breast cancer risk would also be of interest. We recognize that observational studies conducted in free-living populations can examine only the limited range of dietary intake patterns seen in these populations and that they cannot identify which dietary associations are causal. Intervention studies, such as those described below, may be necessary to examine complex dietary patterns and to explore the potential causal nature of the associations we have identified. Ongoing randomized trials that examine effects of lowering fat intake28 or increasing intake of fruits and vegetables27 after breast cancer diagnosis will make important contributions to understanding the association of diet with breast cancer outcomes and the prognostic impact of changing diet after diagnosis, particularly if they yield positive results. These trials should also provide important information on prognostic effects of extremes in intake. The Women’s Intervention Nutrition Study, which will examine the impact of dietary fat reduction on breast cancer outcomes, has reported that women in the control group (whose diet was not modified) consumed 31.5% ± 2.6% of their calories as fat, whereas women in the intervention arm consumed 20.3% ± 2.4% of calories as fat.28 Both of these means fall within the lowest quintile of fat intake in our study, thus extending the range fat intake we were able to investigate in our study. However, if there truly is a significant nonlinear association of diet with breast cancer outcomes, it is possible that these studies will yield false-negative results if they focus on the low extremes in intake and do not include women with intermediate levels of intake. Our study subjects were enrolled between 1989 and 1996. Since enrollment began, rates of overweight and obesity increased in both Canada29,30 and the United States.31,32 Similar trends have been reported for total caloric intake.33 During this time, total fat intake has decreased somewhat,34,35 and there has been a shift towards fast food sources for fat intake. Thus it is possible that women diagnosed with breast cancer today may have diets, and body size, that lie more toward the extremes than the patients we studied. The increase in obesity would be expected to have an adverse effect on breast cancer prognosis; changes in fat intake would have only modest effects overall, but combined with concurrent changes in other nutrients, might also impact breast cancer outcomes. Overall, our research would suggest that diets that minimize extremes in nutrient intake and a lifestyle that results in a normal BMI may be associated with the best breast cancer outcomes.
This research was funded by the Canadian Breast Cancer Research Initiative (grants 6301, 9045, and 12093) and the Medical Research Council of Canada (currently Canadian Institutes of Health Research).
1. Willett WC: Diet and breast cancer. J Intern Med 249:395–411, 2001[CrossRef][Medline] 2. Meyskens FL Jr, Jungi WF, Gerber M, et al: WHO consensus conference on diet and cancer: Members of the breast cancer panel. Eur J Cancer Prev 9:213–216, 2000[CrossRef][Medline] 3. Boyd NF, Lockwood GA, Greenberg CV, et al: Effects of a low-fat high-carbohydrate diet on plasma sex hormones in premenopausal women: Results from a randomized controlled trial—Canadian Diet and Breast Cancer Prevention Study Group. Br J Cancer 76:127–135, 1997[Medline] 4. The Women’s Health Initiative Study Group: Design of the Women’s Health Initiative clinical trial and observational study. Control Clin Trials 19:61–109, 1998[CrossRef][Medline] 5. Henderson MM, Kushi LH, Thompson DJ, et al: Feasibility of a randomized trial of a low-fat diet for the prevention of breast cancer: Dietary compliance in the Women’s Health Trial Vanguard Study. Prev Med 19:115–133, 1990[CrossRef][Medline] 6. Gregorio DI, Emrich LJ, Graham S, et al: Dietary fat consumption and survival among women with breast cancer. J Natl Cancer Inst 75:37–41, 1985[Medline]
7. Newman SC, Miller AB, Howe GR: A study of the effect of weight and dietary fat on breast cancer survival time. Am J Epidemiol 123:767–774, 1986 8. Ewertz M, Gillanders S, Meyer L, et al: Survival of breast cancer patients in relation to factors which affect the risk of developing breast cancer. Int J Cancer 49:526–530, 1991[Medline] 9. Nomura AM, Marchand LL, Kolonel LN, et al: The effect of dietary fat on breast cancer survival among Caucasian and Japanese women in Hawaii. Breast Cancer Res Treat 1:S135–S141, 1991 (suppl 18) 10. Kyogoku S, Hirohata T, Nomura Y, et al: Diet and prognosis of breast cancer. Nutr Cancer 17:271–277, 1992[Medline] 11. Rohan TE, Hiller JE, McMichael AJ: Dietary factors and survival from breast cancer. Nutr Cancer 20:167–177, 1993[Medline]
12. Holm LE, Nordevang E, Hjalmar ML, et al: Treatment failure and dietary habits in women with breast cancer. J Natl Cancer Inst 85:32–36, 1993
13. Jain M, Miller AB, To T: Premorbid diet and the prognosis of women with breast cancer. J Natl Cancer Inst 86:1390–1397, 1994 14. Ingram D: Diet and subsequent survival in women with breast cancer. Br J Cancer 69:592–595, 1994[Medline] 15. Zhang S, Folsom AR, Sellers TA, et al: Better breast cancer survival for postmenopausal women who are less overweight and eat less fat: The Iowa Women’s Health Study. Cancer 76:275–283, 1995[CrossRef][Medline] 16. Hebert JR, Hurley TG, Ma Y: The effect of dietary exposures on recurrence and mortality in early stage breast cancer. Breast Cancer Res Treat 51:17–28, 1998[CrossRef][Medline] 17. Saxe GA, Rock CL, Wicha MS, et al: Diet and risk for breast cancer recurrence and survival. Breast Cancer Res Treat 53:241–253, 1999[CrossRef][Medline] 18. Holmes MD, Stampfer MJ, Colditz GA, et al: Dietary factors and the survival of women with breast carcinoma. Cancer 86:826–835, 1999[CrossRef][Medline]
19. Willett W, Stampfer MJ: Total energy intake: Implications for epidemiologic analyses. Am J Epidemiol 124:17–27, 1986 20. Marshall JR: Diet and breast carcinoma survival: An abundance of hope, a dearth of evidence. Cancer 86:751–753, 1999[CrossRef][Medline]
21. Goodwin PJ, Ennis M, Pritchard KI, et al: Fasting insulin and outcome in early-stage breast cancer: Results of a prospective cohort study. J Clin Oncol 20:42–51, 2002
22. Suissa S, Pollak M, Spitzer WO, et al: Body size and breast cancer prognosis: A statistical explanation of the discrepancies. Cancer Res 49:3113–3116, 1989
23. Calle EE, Thun MJ, Petrelli JM, et al: Body-mass index and mortality in a prospective cohort of U.S. adults. N Engl J Med 341:1097–1105, 1999
24. Block G, Hartman AM, Dresser CM, et al: A data-based approach to diet questionnaire design and testing. Am J Epidemiol 124:453–469, 1986
25. Willett WC, Howe GR, Kushi LH: Adjustment for total energy intake in epidemiologic studies. Am J Clin Nutr 65:1220S–1228S, 1997 26. Johnson RK, Kennedy E: The 2000 Dietary Guidelines for Americans: What are the changes and why were they made? The Dietary Guidelines Advisory Committee. J Am Diet Assoc 100:769–774, 2000[CrossRef][Medline] 27. Pierce JP, Faerber S, Wright FA, et al: Feasibility of a randomized trial of a high-vegetable diet to prevent breast cancer recurrence. Nutr Cancer 28:282–288, 1997[Medline]
28. Chlebowski RT, Blackburn GL, Buzzard IM, et al: Adherence to a dietary fat intake reduction program in postmenopausal women receiving therapy for early breast cancer: The Women’s Intervention Nutrition Study. J Clin Oncol 11:2072–2080, 1993
29. Katzmarzyk PT: The Canadian obesity epidemic, 1985–1998. CMAJ 166:1039–1040, 2002 30. Tremblay MS, Katzmarzyk PT, Willms JD: Temporal trends in overweight and obesity in Canada, 1981–1996. Int J Obes Relat Metab Disord 26:538–543, 2002[CrossRef][Medline]
31. Nelson DE, Bland S, Powell-Griner, et al: State trends in health risk factors and receipt of clinical preventive services among US adults during the 1990’s. JAMA 287:2659–2667, 2002
32. Mokdad AH, Serdula MK, Dietz WH, et al: The spread of the obesity epidemic in the United States, 1991–1998. JAMA 282:1519–1522, 1999 33. Nielsen SJ, Siega-Riz AM, Popkin BM: Trends in energy intake in U. S. between 1997 and 1996: Similar shifts seen across age groups. Obes Res 10:370–378, 2002[Medline] 34. Gray-Donald K, Jacobs-Starkey L, Johnson-Down L: Food habits of Canadians: Reduction in intake over a generation. Can J Public Health 91:381–385, 2000[Medline] 35. Popkin BM, Siega-Riz AM, Haines PS, et al: Where’s the fat? Trends in U. S. diets 1965–1996. Prev Med 32:245–254, 2001[CrossRef][Medline] Submitted June 19, 2002; accepted April 11, 2003. This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
