点击显示 收起
1 From the Department of Epidemiology and Population Health, Albert Einstein College of Medicine, Bronx, NY (SNR); the Departments of Epidemiology and Nutrition (EBR, WCW, and FBH) and Biostatistics (BR), Harvard School of Public Health, Boston, MA; and the Channing Laboratory, Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA (EBR, WCW, and FBH)
2 Supported by NIH grants DK58845 and CA87969. FBH is partly supported by an American Heart Association Established Investigator Award. 3 Reprints not available. Address correspondence to S Rajpathak, Department of Epidemiology and Population Health, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Belfer 1308D, Bronx, NY 10461. E-mail: srajpath{at}aecom.yu.edu.
ABSTRACT
Background: The role of calcium in the maintenance of body weight remains controversial.
Objective: We investigated the association between calcium and dairy intakes and 12-y weight change in US men.
Design: This study was conducted with the use of data from the Health Professionals Follow-up Study, a prospective cohort of men aged 40–75 y in 1986. Data on lifestyle factors and diet were updated biennially with self-administered questionnaires. The participants reported their body weight in 1986 and in 1998. The outcome in our study was 12-y weight change. We used multivariate linear regression to examine how baseline calcium intake (n = 23 504) and change in calcium intake (n = 19 615) were associated with weight change. Because dairy foods are the predominant source of calcium in the diet, we also evaluated a similar association with dairy intake.
Results: In a multivariate analysis with adjustment for potential confounders, baseline or change in intake of total calcium was not significantly associated with weight change. In addition, we did not find any association with dietary, dairy, or supplemental calcium intake when evaluated separately. The men with the largest increase in total dairy intake gained slightly more weight than did the men who decreased intake the most (3.14 compared with 2.57 kg; P for trend = 0.001). This association was primarily due to an increase in high-fat dairy intake. Low-fat dairy intake was not significantly associated with weight change.
Conclusion: Our data do not support the hypothesis that an increase in calcium intake or dairy consumption is associated with lower long-term weight gain in men.
Key Words: Calcium • dairy intake • weight change
INTRODUCTION
Obesity has reached epidemic proportions in the United States. Increased caloric intake (1), in part related to large food portion sizes (2), is thought to have played a major role in this epidemic. Traditionally, research involving energy balance and regulation has been focused on the metabolic fuels, ie, fat, carbohydrate, and protein. However, suboptimal micronutrient nutrition may also contribute to disturbances in energy metabolism. Calcium is an essential macromineral, and its intracellular concentration in adipocytes regulates lipogenesis (3). It was suggested that a low calcium intake stimulates dihydroxyvitamin D and parathyroid hormone secretion and that these calciotrophic hormones cause increased calcium uptake (4). The rise in intracellular calcium promotes lipogenesis and reduces lipolysis, resulting in an increase in fat tissue mass because of lipid filling and adipocyte hypertrophy (5). Therefore, a higher calcium intake could potentially have an antiobesity effect. However, the role of calcium in the maintenance of body weight remains controversial, and only a few studies have examined the relation between calcium intake and long-term weight change. Therefore, we investigated the association between calcium intake and 12-y weight change in the Health Professionals Follow-up Study. The large sample size and the repeated dietary assessment in this cohort allowed us to evaluate weight change with respect to different sources of calcium intake (eg, dietary, supplemental, or dairy) separately.
SUBJECTS AND METHODS
Health Professionals Follow-up Study cohort
The Health Professionals Follow-up Study is a prospective cohort study of 51 529 men aged 40–75 y initiated in 1986 to evaluate the role of diet in chronic diseases. The cohort includes 29 683 dentists, 10 098 veterinarians, 4185 pharmacists, 3745 optometrists, 2218 osteopathic physicians, and 1600 podiatrists. At baseline, all participants completed a mailed questionnaire concerning their diet and medical history. Men without a completed food-frequency questionnaire (FFQ) at baseline, those with unreasonable caloric intakes (>4200 or <800 kcal/d), and those who had left a large number of items blank (>70 items) were excluded from the study (Figure 1). A follow-up questionnaire is sent every 2 y to obtain updated information on incident medical conditions. In addition, information on the participants' dietary intake was updated every 4 y. Physical activity data were expressed as metabolic equivalent hours (MET-h) on the basis of self-reported types and durations of activities during the previous year. One MET-h is equivalent to the energy expended sitting quietly for 1 h (6). The reproducibility and validity of the dietary and physical activity data have been reported in detail elsewhere (7–9). In the present study, we also excluded participants who had a history of chronic disease (cancer other than nonmelanoma skin cancer, cardiovascular disease, and diabetes) at baseline or during follow-up before 1998 (Figure 1) because these diagnoses may affect body weight and dietary intake; we included these participants in a secondary analysis. We conducted the baseline analysis on 23 504 men and the change of intake analysis on 19 615 men. The present study was approved by the Human Subjects Committee at the Harvard School of Public Health.
FIGURE 1.. Participant exclusions and sample sizes for the baseline and change analyses. CVD, cardiovascular disease.
Assessment of nutrient intake
Information on the average frequency of consumption of selected foods and beverages during the preceding year were obtained through semiquantitative FFQs. The participants chose from 9 possible responses that ranged from never to 6 servings/d. The FFQ has a standard portion size alongside the frequency of intake; hence, each person chooses the consumption frequency for the same serving size. We calculated nutrient intake by multiplying the frequency of consumption with the nutrient content and then adding the contributions from all foods and beverages. The food-composition database used to calculate the nutrient values is based primarily on US Department of Agriculture data (10) and was supplemented with the manufacturers' data. We collected additional information on the use of calcium and multivitamin supplements (including dose and frequency of use) from the participants. Total calcium intake was calculated as the sum of dietary and all supplemental intakes. All nutrient intakes (except supplemental calcium and alcohol) were energy-adjusted by the residual method (11). To determine dairy calcium intake, we summed the calcium intake from whole milk, skim milk, low-fat milk, ice cream, cheese, yogurt, and cream consumption. The Pearson correlation coefficient between calcium intake from the FFQ and the average intake of two 1-wk diet records from a subsample of 127 men from this study population was 0.53 (7). In the present study, we evaluated the role of change in calcium intake (1986–1998) as well as the role of baseline calcium intake (in 1986) on weight change. We calculated glycemic load (GL) by multiplying the carbohydrate content of each food by its glycemic index, then multiplied this value by the frequency of consumption and summed the values from all foods. Each unit of GL represents the equivalent of 1 g carbohydrate from white bread. In a validation study, a higher GL was strongly associated with increased triacylglycerol concentrations (12). In addition, GL was associated with an increased risk of diabetes in this cohort (13). Intake of trans-fat was estimated from the FFQ with the use of composition values that were based on analyses performed by Enig et al (14) and Slover et al (15). To assess the validity of our measure of trans-fat intake, the calculated intake was compared with its concentration in subcutaneous adipose tissue in 118 male participants of this cohort with the use of gas-liquid chromatography (16). The Spearman rank correlation between calculated intake as a proportion of fat and the proportion in subcutaneous adipose tissue was 0.34 (P < 0.001). To assess the reproducibility of the measurement of trans-fat intake, the same group of men was asked to complete a second FFQ 1 y after the first; the correlation between the 2 measurements was 0.63. The calculation and validity of trans-fat intake and its health effects in this cohort has been reported previously (16–18).
Outcome assessment
Weight change was defined as the difference between the 1986 and 1998 self-reported body weights. We previously evaluated the reproducibility and validity of self-reported body weight in a subset of the cohort (19). Self-reported weight and the average of 2 technician measures were highly correlated (Pearson's correlation coefficient = 0.97), and there was no significant linear trend in the accuracy of reported weight across quartiles of age or body mass index (BMI).
Statistical analysis
All statistical analyses were performed with SAS software version 8 (Cary, NC). We used age-adjusted and multivariate linear regression to examine the association between 12-y weight change (in kg) and baseline and change in calcium intake. In all analyses, we modeled the main exposure variable in quintiles to avoid a linearity assumption and to reduce the effect of outliers. To evaluate the association of dairy and dietary calcium intake with weight change, we excluded men who reported the use of calcium supplements anytime during the 12-y period (n = 6255). We calculated the least-squares means for changes in body weight across quintiles of baseline calcium intake or change in calcium intake after adjustment for age (<50, 50–54, 55–59, 60–64, or 65 y), smoking (past, current, or never), alcohol consumption (nondrinker, 0.1–4.9, 5.0–9.9, 10.0–14.9, or 15.0 g/d), physical activity (MET-h/d), caloric intake (kcal/d), GL, and intakes of caffeine (mg/d), cereal fiber (g/d), whole grains (g/d), fruit and vegetables (servings/d), soft drinks (servings/d), and trans fat (% of calories). We included these nutrients and food variables in the multivariate models either because they are likely to be associated with a healthy lifestyle or because they were shown to predict weight gain in the present cohort or in other cohorts (20–23).
In the analysis of change in calcium intake, we controlled for both change and baseline levels of covariates, including baseline calcium intake because it was a potential confounder. Because weight change can be dependent on attained weight at baseline, we adjusted for this variable in all models. Such adjustment can also control for bias due to regression to the mean (24). We conducted stratified analyses to evaluate a potential effect modification by age (<65 or 65 y) and intakes of vitamin D and protein using median intake as the cutoffs. Because the predominant source of dietary calcium intake was from dairy product consumption, we conducted similar analyses to evaluate whether baseline or change in dairy product intake (servings/d) was associated with weight change. For all analyses of dietary and dairy intake, we excluded the men who reported calcium supplement use to avoid confounding by supplement use. Tests of linear trend across quintiles of calcium or dairy intake were performed by assigning the median values for the categories and fitting this continuous variable in the model. All statistical tests conducted were two-sided with a type I error () of 0.05, and P values < 0.05 were considered statistically significant.
RESULTS
The mean 12-y (1986–1998) weight gain was 3 kg. During this time, the median intake of total calcium increased from 791 to 899 mg/d; the median dietary calcium intake increased from 736 to 765 mg/d, and calcium supplement use increased from 16.1% to 22.5%. The age-standardized characteristics are shown in Table 1 by quintiles of total calcium intake at baseline. Men in the highest quintile were more physically active, less likely to smoke, and more likely to use calcium supplements than men in the other quintiles. In addition, a higher intake of calcium was associated with lower alcohol and caffeine intakes and a higher intake of vitamin D.
View this table:
TABLE 1. Age-standardized baseline characteristics by quintiles of total calcium intake1
In our study, the men with high dairy intakes were less likely to have hypertension or elevated cholesterol than were the men with low dairy intakes, and high diary intake was associated with low intake of caffeine and high intake of saturated fat, cereal fiber, and vitamin D (data not shown).
Baseline intake analyses
Mean weight gain did not differ significantly by quintile of total, dietary, dairy, or supplemental calcium intake assessed at baseline in 1986 (Table 2). Similarly, we did not find a significant association between total dairy intake and weight gain (Table 3). In the multivariate model, the difference in mean weight gain between extreme quintiles of high-fat dairy intake was small (3.24 kg for the highest quintile compared with 2.86 kg for the lowest quintile; P for trend = 0.03), which indicated that the participants who consumed more high-fat dairy at baseline had a slightly lower 12-y weight gain than the participants who consumed less high-fat dairy. Although this difference for low-fat dairy intake between extreme quintiles was similar (2.83 compared with 3.18 kg), it showed a reverse trend that was not significant (P = 0.17).
View this table:
TABLE 2. Mean 12-y weight change according to quintiles of baseline calcium intake
View this table:
TABLE 3. Mean 12-y weight change according to quintiles of baseline intake of dairy products1
Change in intake analyses
In age-adjusted analyses (Table 4), the men who increased their intake of total calcium the most (quintile 5) had slightly less weight gain than those who most decreased their calcium intake (quintile 1; 2.53 compared with 2.94 kg, respectively; P for trend < 0.001). However, the difference between the 2 groups disappeared after adjustment for potential confounders (2.75 compared with 2.70 kg; P for trend = 0.93). Similarly, dietary, dairy, and supplemental calcium intakes were not significantly associated with weight change in multivariate analyses. The men with the largest increase in total dairy intake gained more weight than did the men who decreased intake the most (3.14 compared with 2.57 kg; P for trend = 0.001), which was primarily attributable to an increase in high-fat dairy intake (3.27 compared with 2.70 kg; P for trend < 0.001) (Table 5). However, we did not find any significant association with change in low-fat dairy intake.
View this table:
TABLE 4. Mean 12-y weight change according to quintiles of change in calcium intake
View this table:
TABLE 5. Mean 12-y weight change according to quintiles of change in intake of dairy products1
In all multivariable analyses, we adjusted for nutrients that varied in intakes across quintiles of calcium or dairy intake (eg, caffeine, cereal fiber, etc). However, addition of these variables in the models had no appreciable effect on our results. In both the baseline and change analyses, we did not find an effect modification by age, smoking, BMI, or intakes of vitamin D, alcohol, or protein. Twelve years is a long time period to evaluate weight change with respect to calcium intake. Therefore, we conducted secondary analyses to evaluate 4- and 8-y weight changes, but did not detect any benefits of calcium or dairy intake. In all analyses of supplemental, dietary, or dairy calcium intakes, additional adjustment for total calcium intake had no appreciable effect on our results. When we performed the analyses without excluding the men who developed a chronic disease during the follow-up period, the results were similar. In addition, the exclusion of the men who reported an intentional weight loss of 4.5 kg (10 lbs) during the follow-up (based on a supplementary questionnaire in 1992; n = 2619) did not change our results.
DISCUSSION
In the present longitudinal study, we found no appreciable association between baseline or change in total calcium intake and a 12-y weight change in this population of healthy middle-aged men. This finding also applies to dairy and nondairy calcium intakes. Total and dietary calcium intakes were associated with lower weight gain in age-adjusted models in the change analysis; however, this association was not observed in multivariate models, which suggests that other diet and lifestyle factors may explain the observed association. Although this confounding was not attributable to any one covariate in particular, the important factors included physical activity, smoking, trans fat intake, and body weight at baseline. In addition, a higher intake of dairy products was not associated with less weight gain during the 12-y follow-up. The modest positive association between dairy intake and weight gain is probably due to components of dairy other than calcium.
The potential effect of calcium intake on body weight was first suggested by Zemel et al (4) in a clinical trial that investigated the antihypertensive effect of calcium in obese African-Americans. In that study, the investigators reported that increasing calcium intake from 400 to 1000 mg/d (2 cups of yogurt/d) for 1 y resulted in a 4.9 kg reduction in body fat. A subsequent review of multiple clinical trials that retrospectively evaluated the role of calcium or dairy on body weight concluded that there was not enough evidence to support the benefits of calcium or dairy intake on weight loss (25). However, most of the studies in the review were not designed or powered to study weight change as the outcome variable. Two recent weight-loss trials evaluated the role of calcium or dairy supplementation on weight as the primary outcome. In a 24-wk randomized clinical trial of 32 obese adults (27 women and 5 men) who were maintained on a 500 kcal/d deficit diet for 6 mo, the patients who received a standard diet (400–500 mg dietary calcium/d) lost a mean (±SD) 6.4 ± 2.5% of their body weight, those who received a high-calcium diet (standard diet supplemented with 800 mg calcium/d) lost 8.6 ± 1.1%, and those who received a high-dairy diet (1200–1300 mg dietary calcium/d) lost 10.9 ± 1.6% (P < 0.01) (26). Thus, that small trial suggested that dietary calcium intake may enhance weight loss secondary to caloric restriction, and this effect was greater for dairy product intake. However, in another 25-wk double-blind, randomized clinical trial on weight loss in 100 overweight and obese women, Shapses et al (27) reported no effect of 1 g calcium supplementation/d on body weight (
In our study, we examined both baseline and changes in calcium and diary intake on weight change. Our analysis suggested that the men who consumed more high-fat dairy products at baseline had slightly less weight gain over time than did the men who consumed less high-fat dairy products. However, high-fat dairy intake decreased over time, on average, in the whole population, and we found that the participants who reduced their dairy intake had less weight gain than did those who increased their intake (Table 5). In addition, this finding may be simply due to chance. Although we found that an increase in total and high-fat dairy intake was associated with greater weight gain, we did not detect a similar association with low-fat dairy intake. This suggests that the weight change associated with increased dairy intake may represent changes in other dietary and lifestyle factors. It is also possible that any weight-reducing effect of calcium obtained from dairy products is compensated for by the caloric content of the dairy products. Some may argue that adjusting for caloric intake in all statistical models may be an overadjustment. However, excluding calories from all models or using non-energy-adjusted calcium intake yielded similar results (data not shown). Dairy foods contain trans fat that may have detrimental health effects; however, in this cohort, dairy intake accounted for <1% of the total trans fat intake. In addition, we adjusted for trans-fat intake in the multivariable analysis.
Because of the observational nature of the present study, our findings need to be interpreted with caution. A potential limitation of our study is that changes in body weight may have been due to changes in either fat mass or lean mass, but we could not differentiate between these 2 components. However, any increase in weight is most likely due to an increase in adiposity in middle-aged men, because lean body mass declines with age (37). Also, measurement error related to dietary assessment may bias the results toward the null and lead to wider CIs of the estimates. However, in our study, the 95% CIs were relatively narrow. In addition, we had the statistical power to detect small differences in weight change because of the large sample size of our study population. Residual confounding is another potential concern, but such confounding is unlikely to explain our results because most age and multivariate associations were similar and we found comparable results for dietary, supplemental, and dairy calcium intakes. Note that in our previous analyses of the same cohort, physical activity, whole grain intake, cereal fiber intake, trans-fat intake, and smoking cessation were significantly associated with weight change (38, 39).
In conclusion, our study does not support the hypothesis that increasing calcium or dairy consumption is associated with lower long-term weight gain in men. Whether calcium supplementation or increased dairy intake is beneficial in preventing weight gain needs to be further studied in long-term randomized trials.
ACKNOWLEDGMENTS
SNR, EBR, WCW, and FBH contributed to the study design, analysis, and the writing of the manuscript. WCW, EBR, and FBH were also responsible for data collection and obtaining funding. FBH supervised the study. BR participated in the analysis and manuscript writing. None of the authors had any conflicts of interest.
REFERENCES