Submitted to: American Society for Bone and Mineral Research
Publication Type: Abstract Only
Publication Acceptance Date: 6/1/2003
Publication Date: 9/1/2003
Citation: TUCKER, K.L., TROY, L., MORITA, K., CUPPLES, A.L., HANNAN, M.T., KIEL, D.P. CARBONATED BEVERAGES COMSUMPTION AND BONE MINERAL DENSITY. AMERICAN SOCIETY FOR BONE AND MINERAL RESEARCH. 2003;18(Suppl 2):S241. Interpretive Summary:
Technical Abstract: Soft drink consumption has been thought to have negative effects on BMD, but studies have shown mixed results. Carbonated soft drinks often displace milk in the diet and introduce phosphoric acid (H3PO4) without calcium. Since the phosphorus content of regular cola is 44-62 mg, and of diet cola 27-39 mg, per 12 oz serving, while most other carbonated beverages contain no phosphorus, we hypothesize that consumption of these specific soft drinks would be associated with lower BMD in adult participants in the Framingham Offspring Study. Valid dietary data and BMD measurements were collected for 1672 women and 1148 men from 1996 to 2001. BMD was measured at the spine and 3 hip sites using a Lunar DPX-L. Dietary intake was assessed with a semi-quantitative food frequency questionnaire that specifically queried respondents on the number of servings of cola and other carbonated beverages consumed, and differentiated between regular, caffeine-free and diet beverages. We regressed each BMD measure onto various measures of soft drink consumption, adjusting for BMI, height, age, energy intake, physical activity score, smoking, alcohol use, use of osteoporosis medication, use of calcium or vitamin D supplements, intake of calcium and vitamin D from diet and, for women, menopause status and estrogen use. There were no significant relationships between total carbonated beverage consumption or non-cola carbonated beverage consumption and BMD at any site. Women, but not men, consuming more than three servings of cola (all types)/d had significantly lower BMD at each of the three hip sites, relative to those consuming less than one serving/d: 2.3% lower at the trochanter (p=0.05), 3.3% lower at the femoral neck (P<0.001), and 5.1% lower at Ward's area (P<0.0005). For the spine those with the highest cola intake had BMD 1.2% lower than those consuming less than one serving/d, but this was not significant. This same pattern of results was generally seen for non-caffeinated and diet cola. Adjusting for caffeine did not change results and there was no significant interaction with calcium intake. These results suggest that cola, but not other carbonated soft drink consumption, contributes to lower BMD in adult women, after adjusting for calcium and caffeine intake. Because similar results were seen with diet and non-caffeinated cola, these associations may be due to the phosphoric acid content of cola.