In this survey, we observed inverse associations between habitual dietary calcium intake and body composition variables in women after adjustment for age, physical activity, energy intake, smoking, alcohol use and calcium supplement use. We also observed that the risk of abdominal obesity was significantly decreased with the increase of each quartile of dietary calcium in multivariable logistic regression models in women after adjustment for the same potential confounders. Previous observational studies have demonstrated inverse relationships between dietary calcium intake and BMI, body fat mass or weight change in Western populations 
. In the America National Health and Nutrition Examination Survey NHANES III data set, after controlling for energy intake, an inverse relationship was observed in women (n
380; P<0.0009) 
Different mechanisms were proposed to explain the effect of dietary calcium on body weight changes. Intracellular calcium ([Ca2+
) is a key regulator of lipid metabolism. Elevated intracellular calcium concentrations stimulate the expression and activity of lipogenic enzymes and reduce lipolysis with a subsequent increased accumulation of fat in adipocytes 
. It is well known that lower dietary calcium intake can lead to increased concentrations of 1,25-dihydroxyvitamin D, favoring an increase in [Ca2+
, which promotes lipogenesis. Conversely, a high calcium intake results in lower concentrations of 1,25-dihydroxyvitamin D and an increase in lipolysis 
. Another possible explanation is that high levels of calcium can reduce absorption of fat in the gut. Some studies in humans and animals have suggested that calcium increases the excretion of fat, presumably by formation of insoluble calcium fatty acid soaps or by binding of bile acids, resulting in malabsorption of fat 
However, we observed that there was a significant sex-dietary calcium interaction on body composition variables and abdominal obesity in a Chinese population. We found the inverse associations between body composition variables, abdominal obesity and dietary calcium intake in Chinese women, but not in men. There have been previous reports of gender-specific effects of calcium intake on body composition. In a study by Jacqmain et al. 
, after adjustment for confounding variables, no relation was observed for men, but a negative relation was observed for women. Loos et al. 
reported that there were no effects in Black women but a negative relation in Black men, White men and White women. Kamycheva et al. 
found that there was no effect in men but a positive relation for women. There were no consistent patterns emerging on a possible gender-specific effect of calcium intake on body composition.
In our study, though it is hard to explain why high habitual dietary calcium intake has beneficial effects only in women, a possible explanation may be the effects of women's sex hormones. Abrams 
and Heaney 
have recently demonstrated that increases in plasma estrogen concentrations were associated with an increase in intestinal calcium absorption, which may result in significant metabolic changes in women.
Based on the findings from previous studies and the present survey, calcium supplements may have a beneficial effect on body weight. However, many studies indicated that calcium from supplements have few effects on body weight or composition. Shapses et al. 
found that supplementation with 1 g calcium citrate malate or calcium citrate per day during a 25-week weight-loss intervention did not influence body fat or body weight in 100 premenopausal and postmenopausal women. Two systematic reviews 
of randomized trials of calcium supplementation in adults did not support a beneficial effect in the calcium supplementation group.
In certain intervention studies using calcium supplements, the baseline habitual calcium intake was more than 1000 mg/d, especially in the Western population, which may lead to inconsistent conclusions between surveys and interventional studies. We therefore investigated the effect of the calcium supplements on body weight and fat mass in a Chinese population with low calcium intake. In the present study, we collected data on dietary calcium and calcium supplements using a validated FFQ and self-report questionnaire. The data showed that there was no appreciable association between use of calcium supplements and body composition variables and obesity prevalence in Chinese people after adjusting for age, dietary calcium intake, physical activity level, energy intake, smoking and alcohol use, in spite of the average dietary intake of calcium being only 430 mg/d in the 8940 survey subjects. In addition, the lack of a calcium supplement effect on body weight or fat mass suggests that elemental calcium from supplemental products do not have the expected effects in a Western population with high dietary calcium intake or in a Chinese population with low dietary calcium intake.
There are several aspects that may explain why dietary calcium causes more beneficial effects on male weight or fat mass than calcium supplements. First, dietary calcium is ingested as part of a mixed meal with the presence of other nutrients, such as proteins, carbohydrates and lipids, which are absorbed over a period of several hours as the meal is digested and passes through the small bowel, resulting in a relatively long-term effect on serum calcium. However, people usually take elemental calcium from supplements alone in a fasting state, instead of with meals, which contributes to a transient effect compared to dietary calcium. Secondly, a number of other bioactive compounds in the diet may promote calcium absorption and enhance calcium function. It has been reported that, in some studies, the combination of dairy protein and dietary calcium enhanced weight loss more effectively than either compound alone 
. Thirdly, as described above, calcium can reduce the absorption of fat in the gut, which is achieved when fat and calcium co-exist in the intestine, which may explain, at least partly, why dietary calcium can promote fat excretion more effectively. Taken together, the source of calcium from foods rather than elemental calcium from supplements had a profound effect on body composition and obesity.
The inferences presented in this cross-sectional study are limited due to the observational and descriptive nature of this evidence. In this survey, we observed that the frequency distribution of responses on calcium supplement use was mainly concentrated in the past five years (about 80%), which may be due to calcium products becoming increasingly available with increasing propaganda and public awareness of healthcare. If there is any effect of calcium supplements on body composition, it should be readily observed because of the recent and concentrated advertisement for calcium supplements. In addition, all information on calcium supplement use was self-reported, thus there was the potential for misclassification of calcium supplement use.
In conclusion, we observed that habitual dietary calcium intake, but not use of calcium supplements, was inversely associated with body composition and abdominal obesity in Chinese women. It is possible that the beneficial effect of calcium on body composition and obesity could be significant if it is taken with naturally calcium-rich content food. Because of the observational nature of the present study, our findings need to be interpreted with caution. Whether dietary calcium, but not elemental calcium from supplements, has beneficial effects on the maintenance of body composition and the prevention of abdominal obesity in Chinese women needs to be assessed in long-term studies.