- Click here for explanation of classification scheme.

The purpose was to re-evaluate 5 clinical studies to explore associations between calcium intake and body weight in women.Two of the studies were cross-sectional, 2 were longitudinal, and 1 was a randomized control trial.

For the 4 longitudinal studies, researchers included only women for whom they had at least 3 observations over time. 

The researchers excluded women who, while under study, developed illnesses that might impact body weight. 

Recruitment Not indicated

Design The researchers examined data accumulated in 5 studies conducted out of thee Osteoporosis Research Center over a 12-year period. Four of the studies have been published elsewhere. One of the studies is an ongoing randomized trial in which blind has not been broken but entry data were available for cross-sectional analysis. For cross-sectional data, BMI was regressed against calcium intake on entry. For longitudinal data, weight change was regressed agains calcium intake. For the controlled trial, difference in amounts of weight gained or lost during observation between calcium-supplemented and placebo-treated groups was tested against null hypothesis.   

Blinding used (if applicable) Blinding is being used in the ongoing randomized trial (blind has not been broken but entry data were available for cross-sectional analysis).

Intervention (if applicable) NA

Statistical Analysis For cross-sectional data, BMI was regressed against calcium intake on entry using standard statistical procedures and the slope of the relationship was considered the outcome variable. For longitudinal data, weight change was regressed agains calcium intake. For the controlled trial, difference in amounts of weight gained or lost during observation between calcium-supplemented and placebo-treated groups was tested against null hypothesis. of zero difference using Student's t test. Multiple linear regression models wee testing using Crunch 4.04.   

 

Timing of Measurements Timing of measurements for the 5 studies is as follows:

• YWS Study-cohort of 184 healthy women in their early 20s followed for 4 years
• TCD Study-similar cohort of young women who participated in a randomoized controlled trial of calcium supplementation
• Nuns Study-a prospective study of calcium metabolism and bone health at 5-year intervals in a cohort of 191 nuns passing from premenopause to postmenopause
• MBx Study -study of bone dynamics and biochemical markers in a cohort of 75 healthy perimenopausal women observed at 5-month intervals over a 5-year period
• Van Study-a randomized controlled trial of calcium supplementation in 216 elderly women  

Dependent Variables

• Weight, Weight Loss, and BMI-weight and height were assessed on entry and at each visit in virually all studies using a Harpendon stadiometer for height and either a balance beam or electronic platform balance for weight. Subjects wore light indoor clothes without shoes. In subjects without osteoporosis, weight was adjusted for height by using BMI, espressed as kd/meters squared. In older subjects with osteroporosis, weight change during observation or treatment was outcome variable.

Independent Variables

• Calcium Intake-for nonintervention studies, 7-day food diaries were evaluated by registered dietitians using a succession of methods over time. For Nuns study (beginning in 1967), intakes were assessed by hand calculation, referring to USDA Handbook 8 and later Bowes and Church. Computer software was used in the other 4 studies and in the Nuns study once it became available. TWS and MBx studies used NutriPractor. TCD began in 1995 and has used Food Processor. For YWS and MBx (which had 6-month visit intervals), only initial dietary analysis was utilized. For Nuns study (which had 5-year visit intervals), average intake values over observation period was used. Calcium intake was expressed as the calcium to protein ratio becasue this expxlicitly factors in the countervaliing effects of the 2 nutrients and this ratio eliminates most portion size estimation error.

 

 

Initial N: not indicated

Ffinal N: 780 women

Age: Participants were clustered into 3 main age groups: 3rd, 5th, and 8th decades

Ethnicity: not indicated

Location: not indicated 

 

1. Slopes of BMI regressed against dietary calcium to protein ratio was significantly negative for the YWS and TCD studies (n=358 3rd decade women). Pooled slope was -0.186 kg/meteres squared/g (P=0.001). Reported kcal intake was not correlated with entry weight or BMI, and none of the multivariate models was superior to simpler bivariate regression of BMI on calcium to protein ratio. There were substantially fewer participants in the overweight zone for calckum intakes above than below the median. Odds ratio for being overweight for calcium intakes below median was 2.25 (P,0.02), with difference even greater for BMI values >30 (i.e., obesity).
2. Using the 2 longitudinal studies (Nuns and MBx, n=216 middle-aged women), change in weight (kg/year) was regressed against dietary calcium to protein ratio (mg/g). In both studies, slope was negative and invidually of borderline significant (0.15>P>0.05). When both studies were pooled, the slope was highly significantly different from 0 (P=0.008). As with the data from younger women (see #1),  multivariate models that incorporated energy, calcium, and protein intake were not superior to bivariate model using either calcium to protein ratio or just calcium intake alone.   
3. Results of the Van study (randomized controlled trial), both groups lost weight over the course of the nearly 4-year observation period. However, weighti change, weighted for duration in study, was -0.671 kg/year in calcium-supplemented group and -0.325 kg/year in the placebo-control group, for a treatment difference of 0.346 kg/year (P<0.025).   

1. Significant negative correlations between calcium intake and weight were found for all 3 age groups (3rd, 5th, and 8th decades).
2. Odds ratio for being overweight (BMI>26) was 2.25 for young women in lower hald of calcium consumpton of their respective study groups (P>0.02).
3. Compared to placebo, the calcium-treated subjects in the controlled trial exhibited a significant weight loss across nearly 4 years of observation. Estimates of the relationship suggest that a 1000-mg calcium intake difference is associated with an 8-kg difference in mean body weight with calcium intake explaining approximately 3% of variance in body weight.

Government:	NIH
Industry:	National Dairy Council Commodity Group:
University/Hospital:	Creighton University
Not-for-profit	0 Foundation associated with industry:

1. No information was provided about race or ethnicity.

2. Researchers noted (per study results) the importance of maintaining a high calcium intake during weight loss or control attempts.

3. Funding was provided, in part, by the National Dairy Council.