- Email: [email protected]
Validation of a calcium assessment tool in postmenopausal Canadian women
Maturitas 69 (2011) 168–172
Contents lists available at ScienceDirect
Maturitas journal homepage: www.elsevier.com/locate/maturitas
Validation of a calcium assessment tool in postmenopausal Canadian women Ada Hung a,b , Maryam Hamidi a,b , Ekaterina Riazantseva a , Lilian Thompson c , Lianne Tile a,d , George Tomlinson d,f , Brooke Stewart a , Angela M. Cheung a,b,d,e,f,∗ a
Osteoporosis and Women’s Health Programs, University Health Network, Toronto General Hospital, Toronto, Canada Institute of Medical Science, University of Toronto, Toronto, Canada Department of Nutritional Sciences, University of Toronto, Toronto, Canada d Division of General Internal Medicine, University of Toronto, Toronto, Canada e Division of Endocrinology, Department of Medicine, University of Toronto, Toronto, Canada f Clinical Epidemiology Program, Department of Health Policy, Management and Evaluation, University of Toronto, Toronto, Canada b c
a r t i c l e
i n f o
Article history: Received 3 August 2010 Received in revised form 18 February 2011 Accepted 28 February 2011
Keywords: Calcium Postmenopausal women Dietary assessment tool Validation study
a b s t r a c t Adequate calcium intake is important for optimal bone health. Assessing dietary calcium intake, however, is neither widely done nor standardized in North American clinical practices. Objective: Our goal was to validate a calcium assessment tool (CAT), a modified version of the Calcium CalculatorTM , against the 3-day food record. Methods: Data were obtained from 348 participants in the ECKO (Evaluation of the Clinical use of vitamin K supplementation in postmenopausal women with Osteopenia) trial. In this study, CAT data was collected at baseline and 3-day food records (FRs) were collected at baseline and 3 months by trained study coordinators. CAT and 3-day FR data were compared using correlations and Bland–Altman plots. Additionally, receiver operator characteristic (ROC) curves of CAT were constructed to identify subjects with low calcium intake at thresholds of 500 mg/day and 1000 mg/day on the 3-day FR curves. Results: Mean calcium intake values per day were 902 mg for the 3-day FRs and 781 mg for the CAT. The Pearson correlation was 0.57 (95% CI: 0.50–0.64). Areas under the ROC curves at thresholds of 500 and 1000 mg calcium were 0.81 (95% CI: 0.73–0.89) and 0.82 (95% CI: 0.78–0.86), respectively. Conclusions: The CAT is a valid tool for the measurement of dietary calcium intake using cut-off values of 500 mg and 1000 mg in postmenopausal women, even though there is only moderate correlation between the CAT and 3-day FR. This tool may facilitate the determination of whether calcium supplements are needed in the clinical setting. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Calcium is an essential nutrient in maintaining optimal health. It is a structural element in the bone and also involved in nerve
Abbreviations: CAT, calcium assessment tool; ECKO„ Evaluation of the Clinical use of vitamin K supplementation in postmenopausal women with Osteopenia; ROC, receiver operator characteristic; FR, food record; CI, confidence intervals; RDA, recommended daily allowance; BMD, bone mineral density; FFQ, food frequency questionnaires; AUROC, area under the ROC curve; ICC, intra-class correlation coefficient. ∗ Corresponding author at: 200 Elizabeth Street, 7 Eaton North – 221, Toronto, Ontario, M5G 2C4, Canada. Tel.: +1 416 340 4301; fax: +1 416 340 4105. E-mail addresses: [email protected] (A. Hung), [email protected] (M. Hamidi), [email protected] (E. Riazantseva), [email protected] (L. Thompson), [email protected] (L. Tile), [email protected] (G. Tomlinson), [email protected] (B. Stewart), [email protected] (A.M. Cheung). 0378-5122/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.maturitas.2011.02.016
conduction, muscle contraction and blood clotting and therefore, serum and cellular levels of calcium are highly regulated. When calcium needs are not met through dietary sources, calcium is released from the skeleton, leaving the bones less dense and more fragile over time. In combination with vitamin D, adequate calcium intake through diet and supplementation has been shown in randomized controlled trials of postmenopausal women to be useful in maintaining BMD [1,2], decreasing fractures [3,4], and even reducing colorectal cancers and gynaecologic cancers [5,6]. On the other hand very high intakes of calcium may lead to an increased risk of kidney stone formation [2], myocardial infarctions [7], hypercalcemia [8], and decreased absorption of other essential minerals. The Recommended Dietary Allowance (RDA) for calcium are 1000 mg/day for adults 19–50 years of age and 1200 mg/day for adults aged ≥51 and the tolerable upper intake level is set at 2500 mg/day [9]. A recent study shows that while Canadian women under age 51 meet the RDA levels, older women may be at risk of
A. Hung et al. / Maturitas 69 (2011) 168–172
deficiency [10]. Given the narrow window of effectiveness in which calcium is needed to maintain health without causing adverse effects, it is important to have a valid tool to measure an individual’s calcium intake through diet and supplements. Assessing dietary calcium intake, however, is neither widely done nor standardized in North American clinical practices. The food record (FR) is the gold standard for assessing calcium intake; however it is a labour intensive method that requires expertise in the use of nutrition analysis software as well as subject training. By contrast, food frequency questionnaires (FFQ) are a more practical and efficient means of collecting dietary data in the clinical setting. Numerous studies have shown FFQs to be valid tools of calcium intake in different countries [11–20] yet, these studies also identified a need to validate FFQs in each region of use, due to cultural and geographic variations. There are currently no published studies using a validated instrument in Canada, although a Calcium CalculatorTM was created by registered dietitians working for the BC Dairy Foundation in the 1980s [21]. The objective of our study was to assess the validity of a modified FFQ from the Calcium CalculatorTM , called the Calcium Assessment Tool (CAT), by examining its agreement with the 3-day FR method among Canadian postmenopausal women. 2. Materials and methods 2.1. Study population and design The subjects for this validation study were postmenopausal Canadian women participating in an osteoporosis prevention trial (ECKO—Evaluation of the Clinical use of vitamin K supplementation in postmenopausal women with Osteopenia) at the University Health Network (UHN), University of Toronto [22]. Women were recruited through health fairs, advertisements and posters from January 2002 to September 2004 in the Greater Toronto Area. Women were included if they were postmenopausal (at least 1 year after their last menses) and had their lowest BMD T-score (total lumbar spine L1–L4, total hip or femoral neck) between −1.0 and −2.0. Women were excluded if they had osteoporosis, fragility fracture after the age of 40, bone medication use in the past 3 months (bisphosphonates, selective estrogen receptor modulators, hormone replacement therapy or calcitonin); any bisphosphonate use for more than 6 months; known metabolic bone diseases or diseases of the liver, kidney, pancreas, lung or heart; history of active cancer in the past 5 years; or chronic oral steroid or coumadin use; or use of high doses of vitamin A (>10,000 IU/day) or E (>400 IU/day). All women gave informed consent prior to the start of study procedures and the protocol was approved by the UHN and University of Toronto research ethics boards. 2.2. Dietary calcium assessment 2.2.1. Dietary calcium assessment using CAT Participants were interviewed by trained study coordinators at the baseline visit and the CAT was administered. The CAT was based on a short FFQ developed by the BC Dairy Foundation, as part of their 2002 brochure version of the Calcium CalculatorTM which is an interactive educational tool [21,23,24]. This FFQ contains 26 items that are most abundant in calcium and commonly eaten in British Columbia (Sydney Massey, personal communication). Portion sizes are given in common household measurements such as slices, cups, tablespoons, cube size and cans. Each portion size contains at least 50 mg of calcium, with increments of 75, 150, 250 and 300 mg per portion. The 2002 brochure version of the BC Dairy Foundation Calcium CalculatorTM asked individuals what they ate the day before. In our study, we asked individuals what they had
169
consumed in the past 7 days to obtain an average daily intake. Use of calcium supplements was asked separately afterwards. The total amount of calcium (in mg) was divided by 7 to find the average intake per day. 2.2.2. Dietary calcium assessment using 3-day food records The subjects were instructed to record their food intake for three consecutive days (two weekdays and one weekend) in two time periods; once immediately after the baseline visit and again the week prior to the 3-month visit. Subjects recorded the description, time, portion sizes (using common household measurements such as cups, slices and tablespoons) and recipes of the foods eaten. A portion size guide, developed by the Fred Hutchinson Cancer Research Center [25], was used in this study. Completed and returned FRs were entered by trained research assistants into the Food Processor software from Elizabeth Stewart Hands and Associates (ESHA) Research [26] and nutritional contents were analyzed. 2.3. Statistical methods The similarity of the calcium intake values between the baseline and 3-month 3-day FRs was assessed using the intra-class correlation coefficient (ICC) [27] (estimated from one-way random effects ANOVA with patients as the random effect), and a comparison of the means. The Bland–Altman method [28] was used to assess agreement of calcium intake as measured by the FR and CAT. Using this method, we estimated the difference in the mean calcium intake produced by the two methods as well as the 95% limits of agreement on the two calcium intake values calculated for any one individual. We also computed the Pearson correlation between the calcium intake as measured by the FR and CAT. We used ROC curves to assess the ability of the CAT to discriminate between subjects with calcium intake on the FR above and below two important thresholds [29]: below 500 mg/day and below 1000 mg/day on the FR. These two thresholds were chosen based on the distributions of dietary calcium consumption in the population [30], along with RDA levels. For each of these thresholds, a different ROC curve was generated by computing the sensitivity and specificity of a range of cut-offs of the CAT values for classifying calcium intake as high or low. The area under the ROC curve (AUROC) and its confidence intervals were calculated by using the equivalence to the Mann–Whitney statistic. Boxplots were used to present the distributions of CAT values for subjects above and below the thresholds on the 3-day FR. 3. Results Out of 440 postmenopausal women in the ECKO trial, 52 had either incomplete CATs or 3-day FRs or both at baseline. At 3 months, 36 women had missing and 4 had incomplete 3-day FRs. Thus 348 women with complete CATs, and baseline and 3-month 3day FRs are included in this study. The demographic characteristics of the 348 subjects are shown in Table 1. The ICC between the baseline and 3-day FRs was 0.61 and the difference in mean calcium intake on the two occasions was 60 mg/day. Calcium intake from the 3-day FRs ranged from 135 to 2398 mg/day, with a mean of 902 mg/day. Calcium intake, as calculated by the CAT, ranged from 50 to 2486 mg/day, with a mean of 781 mg of calcium per day. There was a moderate correlation between the 3-day FRs and CAT (r = 0.57, 95% CI: 0.50–0.64) (Figs. 1 and 2) but the CAT on average underestimates the total calcium consumed, relative to the 3-day FR. The mean difference found between the 3-day FRs and the CAT was 121 mg (95% CI: 83–159) of calcium per day. The estimated 95% limits of agreement for values computed on the same person by the two techniques were −600 to 841 mg of calcium per day. When we examined the performance
170
A. Hung et al. / Maturitas 69 (2011) 168–172
Table 1 Demographic characteristics of 348 postmenopausal women who completed 3 day food records at baseline and three months follow-up, along with a CAT.
Age (range) Age at Menarche (range) Age of menopause Natural menopause (range) Surgical menopause (range) Years since menopause (SD) BMI (kg/cm2 ) (SD) Ethnicity (number and proportion) Caucasian (%) Asian (%) African Canadian (%) Other (%) Education (number and proportion) High school or less (%) Community College (%) University (%) Post graduate (%) Marital status Single/divorced (%) Married/common law (%) Family osteoporosis history (%) Previous HRT use (%) Discontinued within 1 year (%) Current smoker (%) Previous smoker (%) Alcohol use ≤1 serving /day (%) ≥1 serving /day (%) Bone mineral density (g/cm2 ) Spine L1–L4 (SD) Total hip (SD) Femoral neck (SD) Ultra distal radius (SD) 25-Hydroxyvitamin D (nmol/L) (SD)
Mean 59.3 (40.1–82.3) 12.8 (9–17) 50.2 (32–62) 42.1 (26–55) 10.6 (8.2) 25.6 (4.1) 304 (87.4%) 30 (8.6%) 10 (2.9%) 4 (1.1%) 71 (20.4%) 79 (22.7%) 133 (38.2%) 65 (18.7%) 149 (42.8%) 199 (57.2%) 117 (34.1%) 160 (46.0%) 19 (5.5%) 169 (48.6%) 158 (45.4%) 190 (54.6%) 0.920 (0.079) 0.869 (0.077) 0.712 (0.065) 0.406 (0.050) 78.71 (24.57)
of the CAT for identifying subjects below different thresholds of calcium intake per day based on the 3-day FR, we found that the AUROC was 0.81 (95% CI: 0.73–0.89) (Fig. 3a), with a sensitivity of 0.73 (95% CI: 0.56–0.85) and specificity of 0.79 (95% CI: 0.75–0.84) for a threshold of 500 mg/day. Thirty-three participants were below the threshold of 500 mg/day, while 315 were above the threshold. For the threshold of 1000 mg/day, the AUROC was 0.82 (95% CI: 0.78–0.86), with a sensitivity of 0.71 (95% CI: 0.65–0.77) and specificity of 0.72 (95% CI: 0.63–0.79) (Fig. 3b). 230 and 118 participants were respectively below and above this threshold.
Fig. 2. Bland–Altman plots of values from 3-day food records and CAT.
4. Discussion This is the first large study in Canada to validate a short FFQ to assess calcium intake for clinical use in postmenopausal women. Our study showed moderate correlation between the CAT and the 3-day FR. Although, CAT underestimates the 3-day FRs by 121 mg of calcium intake per day, the magnitude of our underestimation of calcium intake by the CAT falls within the range of underestimations found in other calcium FFQ validation studies (range: 6.7–212 mg) [11,14–16,18,20,31–33]. The AUROC values show that the CAT is promising in ranking postmenopausal women above and below certain thresholds of calcium intake. Our results are similar to other validation studies between FFQs and reference methods [11–20]. While most studies showed correlations between 0.49 and 0.79, Khan et al. [15] and Montomoli et al. [17] showed the highest correlations of 0.84 and 0.9, respectively, likely because their subjects lived in small towns with a culturally homogenous diet. This is in contrast to the dietary diversity found in most metropolitan populations. Similar to results in the Magkos et al. study [16], the Bland–Altman plot in our study shows that the differences increase at higher intakes of calcium, which is expected as variability increases with higher calcium levels. On the other hand, the higher value of the 3-day FRs can be attributed to various factors. First, the 3-day FRs have better capture of the total calcium intake than the CAT which has a limited number of food items. Second, the analysis software used in this study also accounted for calcium intake from foods with partial ingredients from mixed food items, such as sesame seeds from a bagel.
Sensitivity
a
1.0
b 1.0
0.8
0.8
0.6 0.4
AUC=0.81 (95% CI: 0.73-0.88)
0.2
0.6 0.4
AUC=0.82 (95% CI: 0.77-0.86)
0.2 0.0
0.0 0.0 0.2 0.4 0.6 0.8 1.0
1-Specificity
Fig. 1. Plot of values (mg/day) from 3-day food records (average of baseline and 3 months follow-up) versus the values for CAT, with Pearson correlation of 0.57.
Sensitivity
Characteristics
0.0 0.2 0.4 0.6 0.8 1.0
1-Specificity
Fig. 3. The receiver operator characteristic curve identifying values above the (a) 500 mg/ day and (b) 1000 mg/ day thresholds using the CAT.
A. Hung et al. / Maturitas 69 (2011) 168–172
Our study has several strengths. First, the CAT questionnaire reflects Canada’s multicultural population by including a variety of foods containing calcium such as tofu and bok choy. Food questionnaires are typically difficult to adapt from other countries for use in North America, due to variations in the local diet of the culture. For example, Khan et al. [15], showed that small crabs account for 33.8% of the calcium intake of a population in the Red River Delta in Vietnam, however, dairy products contributed only 0.38%. By contrast, dairy products accounted for at least 46% of calcium intake in Canadian diets [34]. Second, our study contains a large cohort of 348 postmenopausal women, compared to other North American studies with typically fewer than 100 postmenopausal women [11,13,19,35–37]. Finally, the CAT enables the clinician to efficiently calculate the calcium intake directly onto one record sheet which may eliminate any potential transcription errors. By contrast, other FFQs commonly use a separate reference sheet to calculate intake values [11,12,15,16,20,35,38,39]. Our study has several limitations. First, despite the availability of multicultural foods in supermarkets and restaurants our subjects were predominantly Caucasian postmenopausal women recruited from one geographic region, the Greater Toronto Area. Second, there are limitations related to our dietary intake instruments. Recall bias may be associated with the CAT, as food intake is assessed over the past 7 days. With 3-day FRs, there may be issues of underreporting and simplification, as study participants are faced with the new and unusual routine of recording every food item eaten throughout the day. However, despite these potential limitations, when compared to 1- and 7-day reference methods, the 3-day FR is a good compromise between reliability, compliance and daily dietary variability. Finally, even though nutritionists were available to clarify any unclear dietary items, misinterpretations may have occurred when entering the food items into the Food Processor software. Despite the limitations we have noted, our study shows that the CAT is a valid tool and can be used to assess dietary calcium intake and adequacy in postmenopausal women when 3-day FRs are not possible. When the CAT is used in the clinical setting, clinicians must be aware that the values obtained may be slightly lower than the patient’s actual intake. Contributors Dr. Angela Cheung led the study concept, design, acquisition of data and critical review of the manuscript and had primary responsibility for final content. Ada Hung actively participated in data management and was involved in the analysis and interpretation of data, drafting of the manuscript, and critical review of the manuscript. Maryam Hamidi contributed to the conception and design of the study, acquired and had full access to all of the data, and was involved in the dietary data management, entry and analysis and critical review of the manuscript. George Tomlinson led the statistical analysis and was involved in the critical review of the manuscript. Ekaterina Riazantseva, Lilian Thompson, Lianne Tile, and Brooke Stewart contributed to data acquisition and critically reviewed the manuscript. Competing interests None. Funding The ECKO trial was supported by a CIHR operating grant. Dr. A.M. Cheung has been supported by a 5-year CIHR mid-career award in postmenopausal osteoporosis
171
Acknowledgments We thank all our subjects for participating in this study. We also thank Lidia Matic for data entry and organization of the food record data, and Yena Ahn and Zoe Agnidis for their assistance in preparation of the manuscript for submission. References [1] Chapuy MC, Arlot ME, Duboeuf F, Brun J, Crouzet B, Arnaud S, et al. Vitamin D3 and calcium to prevent hip fractures in the elderly women. N Engl J Med 1992;327:1637–42. [2] Jackson RD, LaCroix AZ, Gass M, Wallace RB, Robbins J, Lewis CE, et al. Calcium plus vitamin D supplementation and the risk of fractures. N Engl J Med 2006;354:669–83. [3] Dwson-Hughes B, Dallal GE, Krall EA, Sadowski L, Sahyoun N, Tannenbaum S. A controlled trial of the effect of calcium supplementation on bone density in postmenopausal women. N Engl J Med 1990;323:878–83. [4] Reid IR, Ames RW, Evans MC, Gamble GD, Sharpe SJ. Long-term effects of calcium supplementation on bone loss and fractures in postmenopausal women: a randomized controlled trial. Am J Med 1995;98:331–5. [5] Huncharek M, Muscat J, Kupelnick B. Colorectal cancer risk and dietary intake of calcium, vitamin D, and dairy products: a meta-analysis of 26,335 cases from 60 observational studies. Nutr Cancer 2009;61:47–69. [6] McCullough ML, Bandera EV, Moore DF, Kushi LH. Vitamin D and calcium intake in relation to risk of endometrial cancer: a systematic review of the literature. Prev Med 2008;46:298–302. [7] Bolland MJ, Avenell A, Baron JA, Grey A, MacLennan GS, Gamble GD, et al. Effect of calcium supplements on risk of myocardial infarction and cardiovascular events: meta-analysis. BMJ 2010;341:c3691. [8] Patel AM, Goldfarb S. Got calcium? Welcome to the calcium-alkali syndrome. J Am Soc Nephrol 2010;21:1440–3. [9] Institute of Medicine (IOM). DRIs for Calcium and Vitamin D. Institute of Medicine of The National Academies. Available from: http://www.iom.edu/Reports/2010/Dietary-Reference-Intakes-for-Calciumand-Vitamin-D/DRI-Values.aspx [accessed Feb 17, 2011]. [10] Poliquin S, Joseph L, Gray-Donald K. Calcium and vitamin D intakes in an adult Canadian population. Can J Diet Pract Res 2009;70:21–7. [11] Blalock SJ, Norton LL, Patel RA, Cabral K, Thomas CL. Development and assessment of a short instrument for assessing dietary intakes of calcium and vitamin D. J Am Pharm Assoc 2003;43:685–93. [12] Chee WS, Suriah AR, Zaitun Y, Chan SP, Yap SL, Chan YM. Dietary calcium intake in postmenopausal Malaysian women: comparison between the food frequency questionnaire and three-day food records. Asia Pac J Clin Nutr 2002;11:142–6. [13] Cummings SR, Block G, McHenry K, Baron RB. Evaluation of two food frequency methods of measuring dietary calcium intake. Am J Epidemiol 1987;126:796–802. [14] Haines CJ, Chung TK, Leung PC, Leung DH, Wong MY, Lam LL. Dietary calcium intake in postmenopausal Chinese women. Eur J Clin Nutr 1994;48:591–4. [15] Khan NC, Mai lB, Hien VT, Lam NT, Hoa VQ, Phuong TM, et al. Development and validation of food frequency questionnaire to assess calcium intake in postmenopausal Vietnamese women. J Nutr Sci Vitaminol (Tokyo) 2008;54: 124–9. [16] Magkos F, Yannakoulia M. Methodology of dietary assessment in athletes: concepts and pitfalls. Curr Opin Clin Nutr Metab Care 2003;6:539–49. [17] Montomoli M, Gonnelli S, Giacchi M, Mattei R, Cuda C, Rossi S, et al. Validation of a food frequency questionnaire for nutritional calcium intake assessment in Italian women. Eur J Clin Nutr 2002;56:21–30. [18] Sato Y, Tamaki J, Kitayama F, Kusaka Y, Kodera Y, Koutani A, et al. Development of a food-frequency questionnaire to measure the dietary calcium intake of adult Japanese women. Tohoku J Exp Med 2005;207:217–22. [19] Smith BA, Morgan SL, Vaughn WH, Fox L, Canfield GJ, Bartolucci AA. Comparison of a computer-based food frequency questionnaire for calcium intake with 2 other assessment tools. J Am Diet Assoc 1999;99:1579–81. [20] Uenishi K, Ishida H, Nakamura K. Development of a simple food frequency questionnaire to estimate intakes of calcium and other nutrients for the prevention and management of osteoporosis. J Nutr Sci Vitaminol (Tokyo) 2008;54:25–9. [21] BC Dairy Foundation. Calcium Calculator. Available from: http://bcdairyfoundation.ca/interactive/calcium-calculator/ [accessed Feb 17, 2011]. [22] Cheung AM, Tile L, Lee Y, Tomlinson G, Hawker G, Scher J, et al. Vitamin K supplementation in postmenopausal women with osteopenia (ECKO trial): a randomized controlled trial. PLoS Med 2008;5:e196. [23] Osteoporosis Canada. Calcium Calculator. Available from: [accessed http://www.osteoporosis.ca/index.php/ci id/5355/la id/1.htm Feb 17, 2011]. [24] Washington State Dairy Council. Calcium Calculator. Available from: http://www.eatsmart.org/games/c calculator/ [accessed Feb 17, 2011]. [25] Fred Hutchinson Cancer Research Center. Fred Hutchinson Cancer Research Center Food frequency questionnaires. Available from: www.fhcrc.org/science/shared resources/nutrition/ffq/index.html [accessed Feb 17, 2011].
172
A. Hung et al. / Maturitas 69 (2011) 168–172
[26] ESAH Research. The Food Processor. Available from: http://www.esha.com [accessed Feb 17, 2011]. [27] Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–8. [28] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10. [29] Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29–36. [30] Vatanparast H, Dolega-Cieszkowski JH, Whiting SJ. Many adult Canadians are not meeting current calcium recommendations from food and supplement intake. Appl Physiol Nutr Metab 2009;34:191–6. [31] Angbratt M, Moller M. Questionnaire about calcium intake: can we trust the answers? Osteoporos Int 1999;9:220–5. [32] Green JH, Booth CL, Bunning RL. Assessment of a rapid method for assessing adequacy of calcium intake. Asia Pac J Clin Nutr 2002;11:147–50. [33] Pasco JA, Sanders KM, Henry MJ, Nicholson GC, Seeman E, Kotowicz MA. Calcium intakes among Australian women: Geelong Osteoporosis Study. Aust N Z J Med 2000;30:21–7.
[34] Johnson-Down L, Ritter H, Starkey LJ, Gray-Donald K. Primary food sources of nutrients in the diet of Canadian adults. Can J Diet Pract Res 2006;67: 7–13. [35] Brown JL, Griebler R. Reliability of a short and long version of the block food frequency form for assessing changes in calcium intake. J Am Diet Assoc 1993;93:784–9. [36] Musgrave KO, Giambalvo L, Leclerc HL, Cook RA, Rosen CJ. Validation of a quantitative food frequency questionnaire for rapid assessment of dietary calcium intake. J Am Diet Assoc 1989;89:1484–8. [37] Sebring NG, Denkinger BI, Menzie CM, Yanoff LB, Parikh SJ, Yanovski JA. Validation of three food frequency questionnaires to assess dietary calcium intake in adults. J Am Diet Assoc 2007;107:752–9. [38] Angus RM, Sambrook PN, Pocock NA, Eisman JA. A simple method for assessing calcium intake in Caucasian women. J Am Diet Assoc 1989;89: 209–14. [39] Clover E, Miller M, Bannerman E, Magarey A. Relative validation of a short food frequency questionnaire to assess calcium intake in older adults. Aust N Z J Public Health 2007;31:450–8.
Contents lists available at ScienceDirect
Maturitas journal homepage: www.elsevier.com/locate/maturitas
Validation of a calcium assessment tool in postmenopausal Canadian women Ada Hung a,b , Maryam Hamidi a,b , Ekaterina Riazantseva a , Lilian Thompson c , Lianne Tile a,d , George Tomlinson d,f , Brooke Stewart a , Angela M. Cheung a,b,d,e,f,∗ a
Osteoporosis and Women’s Health Programs, University Health Network, Toronto General Hospital, Toronto, Canada Institute of Medical Science, University of Toronto, Toronto, Canada Department of Nutritional Sciences, University of Toronto, Toronto, Canada d Division of General Internal Medicine, University of Toronto, Toronto, Canada e Division of Endocrinology, Department of Medicine, University of Toronto, Toronto, Canada f Clinical Epidemiology Program, Department of Health Policy, Management and Evaluation, University of Toronto, Toronto, Canada b c
a r t i c l e
i n f o
Article history: Received 3 August 2010 Received in revised form 18 February 2011 Accepted 28 February 2011
Keywords: Calcium Postmenopausal women Dietary assessment tool Validation study
a b s t r a c t Adequate calcium intake is important for optimal bone health. Assessing dietary calcium intake, however, is neither widely done nor standardized in North American clinical practices. Objective: Our goal was to validate a calcium assessment tool (CAT), a modified version of the Calcium CalculatorTM , against the 3-day food record. Methods: Data were obtained from 348 participants in the ECKO (Evaluation of the Clinical use of vitamin K supplementation in postmenopausal women with Osteopenia) trial. In this study, CAT data was collected at baseline and 3-day food records (FRs) were collected at baseline and 3 months by trained study coordinators. CAT and 3-day FR data were compared using correlations and Bland–Altman plots. Additionally, receiver operator characteristic (ROC) curves of CAT were constructed to identify subjects with low calcium intake at thresholds of 500 mg/day and 1000 mg/day on the 3-day FR curves. Results: Mean calcium intake values per day were 902 mg for the 3-day FRs and 781 mg for the CAT. The Pearson correlation was 0.57 (95% CI: 0.50–0.64). Areas under the ROC curves at thresholds of 500 and 1000 mg calcium were 0.81 (95% CI: 0.73–0.89) and 0.82 (95% CI: 0.78–0.86), respectively. Conclusions: The CAT is a valid tool for the measurement of dietary calcium intake using cut-off values of 500 mg and 1000 mg in postmenopausal women, even though there is only moderate correlation between the CAT and 3-day FR. This tool may facilitate the determination of whether calcium supplements are needed in the clinical setting. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Calcium is an essential nutrient in maintaining optimal health. It is a structural element in the bone and also involved in nerve
Abbreviations: CAT, calcium assessment tool; ECKO„ Evaluation of the Clinical use of vitamin K supplementation in postmenopausal women with Osteopenia; ROC, receiver operator characteristic; FR, food record; CI, confidence intervals; RDA, recommended daily allowance; BMD, bone mineral density; FFQ, food frequency questionnaires; AUROC, area under the ROC curve; ICC, intra-class correlation coefficient. ∗ Corresponding author at: 200 Elizabeth Street, 7 Eaton North – 221, Toronto, Ontario, M5G 2C4, Canada. Tel.: +1 416 340 4301; fax: +1 416 340 4105. E-mail addresses: [email protected] (A. Hung), [email protected] (M. Hamidi), [email protected] (E. Riazantseva), [email protected] (L. Thompson), [email protected] (L. Tile), [email protected] (G. Tomlinson), [email protected] (B. Stewart), [email protected] (A.M. Cheung). 0378-5122/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.maturitas.2011.02.016
conduction, muscle contraction and blood clotting and therefore, serum and cellular levels of calcium are highly regulated. When calcium needs are not met through dietary sources, calcium is released from the skeleton, leaving the bones less dense and more fragile over time. In combination with vitamin D, adequate calcium intake through diet and supplementation has been shown in randomized controlled trials of postmenopausal women to be useful in maintaining BMD [1,2], decreasing fractures [3,4], and even reducing colorectal cancers and gynaecologic cancers [5,6]. On the other hand very high intakes of calcium may lead to an increased risk of kidney stone formation [2], myocardial infarctions [7], hypercalcemia [8], and decreased absorption of other essential minerals. The Recommended Dietary Allowance (RDA) for calcium are 1000 mg/day for adults 19–50 years of age and 1200 mg/day for adults aged ≥51 and the tolerable upper intake level is set at 2500 mg/day [9]. A recent study shows that while Canadian women under age 51 meet the RDA levels, older women may be at risk of
A. Hung et al. / Maturitas 69 (2011) 168–172
deficiency [10]. Given the narrow window of effectiveness in which calcium is needed to maintain health without causing adverse effects, it is important to have a valid tool to measure an individual’s calcium intake through diet and supplements. Assessing dietary calcium intake, however, is neither widely done nor standardized in North American clinical practices. The food record (FR) is the gold standard for assessing calcium intake; however it is a labour intensive method that requires expertise in the use of nutrition analysis software as well as subject training. By contrast, food frequency questionnaires (FFQ) are a more practical and efficient means of collecting dietary data in the clinical setting. Numerous studies have shown FFQs to be valid tools of calcium intake in different countries [11–20] yet, these studies also identified a need to validate FFQs in each region of use, due to cultural and geographic variations. There are currently no published studies using a validated instrument in Canada, although a Calcium CalculatorTM was created by registered dietitians working for the BC Dairy Foundation in the 1980s [21]. The objective of our study was to assess the validity of a modified FFQ from the Calcium CalculatorTM , called the Calcium Assessment Tool (CAT), by examining its agreement with the 3-day FR method among Canadian postmenopausal women. 2. Materials and methods 2.1. Study population and design The subjects for this validation study were postmenopausal Canadian women participating in an osteoporosis prevention trial (ECKO—Evaluation of the Clinical use of vitamin K supplementation in postmenopausal women with Osteopenia) at the University Health Network (UHN), University of Toronto [22]. Women were recruited through health fairs, advertisements and posters from January 2002 to September 2004 in the Greater Toronto Area. Women were included if they were postmenopausal (at least 1 year after their last menses) and had their lowest BMD T-score (total lumbar spine L1–L4, total hip or femoral neck) between −1.0 and −2.0. Women were excluded if they had osteoporosis, fragility fracture after the age of 40, bone medication use in the past 3 months (bisphosphonates, selective estrogen receptor modulators, hormone replacement therapy or calcitonin); any bisphosphonate use for more than 6 months; known metabolic bone diseases or diseases of the liver, kidney, pancreas, lung or heart; history of active cancer in the past 5 years; or chronic oral steroid or coumadin use; or use of high doses of vitamin A (>10,000 IU/day) or E (>400 IU/day). All women gave informed consent prior to the start of study procedures and the protocol was approved by the UHN and University of Toronto research ethics boards. 2.2. Dietary calcium assessment 2.2.1. Dietary calcium assessment using CAT Participants were interviewed by trained study coordinators at the baseline visit and the CAT was administered. The CAT was based on a short FFQ developed by the BC Dairy Foundation, as part of their 2002 brochure version of the Calcium CalculatorTM which is an interactive educational tool [21,23,24]. This FFQ contains 26 items that are most abundant in calcium and commonly eaten in British Columbia (Sydney Massey, personal communication). Portion sizes are given in common household measurements such as slices, cups, tablespoons, cube size and cans. Each portion size contains at least 50 mg of calcium, with increments of 75, 150, 250 and 300 mg per portion. The 2002 brochure version of the BC Dairy Foundation Calcium CalculatorTM asked individuals what they ate the day before. In our study, we asked individuals what they had
169
consumed in the past 7 days to obtain an average daily intake. Use of calcium supplements was asked separately afterwards. The total amount of calcium (in mg) was divided by 7 to find the average intake per day. 2.2.2. Dietary calcium assessment using 3-day food records The subjects were instructed to record their food intake for three consecutive days (two weekdays and one weekend) in two time periods; once immediately after the baseline visit and again the week prior to the 3-month visit. Subjects recorded the description, time, portion sizes (using common household measurements such as cups, slices and tablespoons) and recipes of the foods eaten. A portion size guide, developed by the Fred Hutchinson Cancer Research Center [25], was used in this study. Completed and returned FRs were entered by trained research assistants into the Food Processor software from Elizabeth Stewart Hands and Associates (ESHA) Research [26] and nutritional contents were analyzed. 2.3. Statistical methods The similarity of the calcium intake values between the baseline and 3-month 3-day FRs was assessed using the intra-class correlation coefficient (ICC) [27] (estimated from one-way random effects ANOVA with patients as the random effect), and a comparison of the means. The Bland–Altman method [28] was used to assess agreement of calcium intake as measured by the FR and CAT. Using this method, we estimated the difference in the mean calcium intake produced by the two methods as well as the 95% limits of agreement on the two calcium intake values calculated for any one individual. We also computed the Pearson correlation between the calcium intake as measured by the FR and CAT. We used ROC curves to assess the ability of the CAT to discriminate between subjects with calcium intake on the FR above and below two important thresholds [29]: below 500 mg/day and below 1000 mg/day on the FR. These two thresholds were chosen based on the distributions of dietary calcium consumption in the population [30], along with RDA levels. For each of these thresholds, a different ROC curve was generated by computing the sensitivity and specificity of a range of cut-offs of the CAT values for classifying calcium intake as high or low. The area under the ROC curve (AUROC) and its confidence intervals were calculated by using the equivalence to the Mann–Whitney statistic. Boxplots were used to present the distributions of CAT values for subjects above and below the thresholds on the 3-day FR. 3. Results Out of 440 postmenopausal women in the ECKO trial, 52 had either incomplete CATs or 3-day FRs or both at baseline. At 3 months, 36 women had missing and 4 had incomplete 3-day FRs. Thus 348 women with complete CATs, and baseline and 3-month 3day FRs are included in this study. The demographic characteristics of the 348 subjects are shown in Table 1. The ICC between the baseline and 3-day FRs was 0.61 and the difference in mean calcium intake on the two occasions was 60 mg/day. Calcium intake from the 3-day FRs ranged from 135 to 2398 mg/day, with a mean of 902 mg/day. Calcium intake, as calculated by the CAT, ranged from 50 to 2486 mg/day, with a mean of 781 mg of calcium per day. There was a moderate correlation between the 3-day FRs and CAT (r = 0.57, 95% CI: 0.50–0.64) (Figs. 1 and 2) but the CAT on average underestimates the total calcium consumed, relative to the 3-day FR. The mean difference found between the 3-day FRs and the CAT was 121 mg (95% CI: 83–159) of calcium per day. The estimated 95% limits of agreement for values computed on the same person by the two techniques were −600 to 841 mg of calcium per day. When we examined the performance
170
A. Hung et al. / Maturitas 69 (2011) 168–172
Table 1 Demographic characteristics of 348 postmenopausal women who completed 3 day food records at baseline and three months follow-up, along with a CAT.
Age (range) Age at Menarche (range) Age of menopause Natural menopause (range) Surgical menopause (range) Years since menopause (SD) BMI (kg/cm2 ) (SD) Ethnicity (number and proportion) Caucasian (%) Asian (%) African Canadian (%) Other (%) Education (number and proportion) High school or less (%) Community College (%) University (%) Post graduate (%) Marital status Single/divorced (%) Married/common law (%) Family osteoporosis history (%) Previous HRT use (%) Discontinued within 1 year (%) Current smoker (%) Previous smoker (%) Alcohol use ≤1 serving /day (%) ≥1 serving /day (%) Bone mineral density (g/cm2 ) Spine L1–L4 (SD) Total hip (SD) Femoral neck (SD) Ultra distal radius (SD) 25-Hydroxyvitamin D (nmol/L) (SD)
Mean 59.3 (40.1–82.3) 12.8 (9–17) 50.2 (32–62) 42.1 (26–55) 10.6 (8.2) 25.6 (4.1) 304 (87.4%) 30 (8.6%) 10 (2.9%) 4 (1.1%) 71 (20.4%) 79 (22.7%) 133 (38.2%) 65 (18.7%) 149 (42.8%) 199 (57.2%) 117 (34.1%) 160 (46.0%) 19 (5.5%) 169 (48.6%) 158 (45.4%) 190 (54.6%) 0.920 (0.079) 0.869 (0.077) 0.712 (0.065) 0.406 (0.050) 78.71 (24.57)
of the CAT for identifying subjects below different thresholds of calcium intake per day based on the 3-day FR, we found that the AUROC was 0.81 (95% CI: 0.73–0.89) (Fig. 3a), with a sensitivity of 0.73 (95% CI: 0.56–0.85) and specificity of 0.79 (95% CI: 0.75–0.84) for a threshold of 500 mg/day. Thirty-three participants were below the threshold of 500 mg/day, while 315 were above the threshold. For the threshold of 1000 mg/day, the AUROC was 0.82 (95% CI: 0.78–0.86), with a sensitivity of 0.71 (95% CI: 0.65–0.77) and specificity of 0.72 (95% CI: 0.63–0.79) (Fig. 3b). 230 and 118 participants were respectively below and above this threshold.
Fig. 2. Bland–Altman plots of values from 3-day food records and CAT.
4. Discussion This is the first large study in Canada to validate a short FFQ to assess calcium intake for clinical use in postmenopausal women. Our study showed moderate correlation between the CAT and the 3-day FR. Although, CAT underestimates the 3-day FRs by 121 mg of calcium intake per day, the magnitude of our underestimation of calcium intake by the CAT falls within the range of underestimations found in other calcium FFQ validation studies (range: 6.7–212 mg) [11,14–16,18,20,31–33]. The AUROC values show that the CAT is promising in ranking postmenopausal women above and below certain thresholds of calcium intake. Our results are similar to other validation studies between FFQs and reference methods [11–20]. While most studies showed correlations between 0.49 and 0.79, Khan et al. [15] and Montomoli et al. [17] showed the highest correlations of 0.84 and 0.9, respectively, likely because their subjects lived in small towns with a culturally homogenous diet. This is in contrast to the dietary diversity found in most metropolitan populations. Similar to results in the Magkos et al. study [16], the Bland–Altman plot in our study shows that the differences increase at higher intakes of calcium, which is expected as variability increases with higher calcium levels. On the other hand, the higher value of the 3-day FRs can be attributed to various factors. First, the 3-day FRs have better capture of the total calcium intake than the CAT which has a limited number of food items. Second, the analysis software used in this study also accounted for calcium intake from foods with partial ingredients from mixed food items, such as sesame seeds from a bagel.
Sensitivity
a
1.0
b 1.0
0.8
0.8
0.6 0.4
AUC=0.81 (95% CI: 0.73-0.88)
0.2
0.6 0.4
AUC=0.82 (95% CI: 0.77-0.86)
0.2 0.0
0.0 0.0 0.2 0.4 0.6 0.8 1.0
1-Specificity
Fig. 1. Plot of values (mg/day) from 3-day food records (average of baseline and 3 months follow-up) versus the values for CAT, with Pearson correlation of 0.57.
Sensitivity
Characteristics
0.0 0.2 0.4 0.6 0.8 1.0
1-Specificity
Fig. 3. The receiver operator characteristic curve identifying values above the (a) 500 mg/ day and (b) 1000 mg/ day thresholds using the CAT.
A. Hung et al. / Maturitas 69 (2011) 168–172
Our study has several strengths. First, the CAT questionnaire reflects Canada’s multicultural population by including a variety of foods containing calcium such as tofu and bok choy. Food questionnaires are typically difficult to adapt from other countries for use in North America, due to variations in the local diet of the culture. For example, Khan et al. [15], showed that small crabs account for 33.8% of the calcium intake of a population in the Red River Delta in Vietnam, however, dairy products contributed only 0.38%. By contrast, dairy products accounted for at least 46% of calcium intake in Canadian diets [34]. Second, our study contains a large cohort of 348 postmenopausal women, compared to other North American studies with typically fewer than 100 postmenopausal women [11,13,19,35–37]. Finally, the CAT enables the clinician to efficiently calculate the calcium intake directly onto one record sheet which may eliminate any potential transcription errors. By contrast, other FFQs commonly use a separate reference sheet to calculate intake values [11,12,15,16,20,35,38,39]. Our study has several limitations. First, despite the availability of multicultural foods in supermarkets and restaurants our subjects were predominantly Caucasian postmenopausal women recruited from one geographic region, the Greater Toronto Area. Second, there are limitations related to our dietary intake instruments. Recall bias may be associated with the CAT, as food intake is assessed over the past 7 days. With 3-day FRs, there may be issues of underreporting and simplification, as study participants are faced with the new and unusual routine of recording every food item eaten throughout the day. However, despite these potential limitations, when compared to 1- and 7-day reference methods, the 3-day FR is a good compromise between reliability, compliance and daily dietary variability. Finally, even though nutritionists were available to clarify any unclear dietary items, misinterpretations may have occurred when entering the food items into the Food Processor software. Despite the limitations we have noted, our study shows that the CAT is a valid tool and can be used to assess dietary calcium intake and adequacy in postmenopausal women when 3-day FRs are not possible. When the CAT is used in the clinical setting, clinicians must be aware that the values obtained may be slightly lower than the patient’s actual intake. Contributors Dr. Angela Cheung led the study concept, design, acquisition of data and critical review of the manuscript and had primary responsibility for final content. Ada Hung actively participated in data management and was involved in the analysis and interpretation of data, drafting of the manuscript, and critical review of the manuscript. Maryam Hamidi contributed to the conception and design of the study, acquired and had full access to all of the data, and was involved in the dietary data management, entry and analysis and critical review of the manuscript. George Tomlinson led the statistical analysis and was involved in the critical review of the manuscript. Ekaterina Riazantseva, Lilian Thompson, Lianne Tile, and Brooke Stewart contributed to data acquisition and critically reviewed the manuscript. Competing interests None. Funding The ECKO trial was supported by a CIHR operating grant. Dr. A.M. Cheung has been supported by a 5-year CIHR mid-career award in postmenopausal osteoporosis
171
Acknowledgments We thank all our subjects for participating in this study. We also thank Lidia Matic for data entry and organization of the food record data, and Yena Ahn and Zoe Agnidis for their assistance in preparation of the manuscript for submission. References [1] Chapuy MC, Arlot ME, Duboeuf F, Brun J, Crouzet B, Arnaud S, et al. Vitamin D3 and calcium to prevent hip fractures in the elderly women. N Engl J Med 1992;327:1637–42. [2] Jackson RD, LaCroix AZ, Gass M, Wallace RB, Robbins J, Lewis CE, et al. Calcium plus vitamin D supplementation and the risk of fractures. N Engl J Med 2006;354:669–83. [3] Dwson-Hughes B, Dallal GE, Krall EA, Sadowski L, Sahyoun N, Tannenbaum S. A controlled trial of the effect of calcium supplementation on bone density in postmenopausal women. N Engl J Med 1990;323:878–83. [4] Reid IR, Ames RW, Evans MC, Gamble GD, Sharpe SJ. Long-term effects of calcium supplementation on bone loss and fractures in postmenopausal women: a randomized controlled trial. Am J Med 1995;98:331–5. [5] Huncharek M, Muscat J, Kupelnick B. Colorectal cancer risk and dietary intake of calcium, vitamin D, and dairy products: a meta-analysis of 26,335 cases from 60 observational studies. Nutr Cancer 2009;61:47–69. [6] McCullough ML, Bandera EV, Moore DF, Kushi LH. Vitamin D and calcium intake in relation to risk of endometrial cancer: a systematic review of the literature. Prev Med 2008;46:298–302. [7] Bolland MJ, Avenell A, Baron JA, Grey A, MacLennan GS, Gamble GD, et al. Effect of calcium supplements on risk of myocardial infarction and cardiovascular events: meta-analysis. BMJ 2010;341:c3691. [8] Patel AM, Goldfarb S. Got calcium? Welcome to the calcium-alkali syndrome. J Am Soc Nephrol 2010;21:1440–3. [9] Institute of Medicine (IOM). DRIs for Calcium and Vitamin D. Institute of Medicine of The National Academies. Available from: http://www.iom.edu/Reports/2010/Dietary-Reference-Intakes-for-Calciumand-Vitamin-D/DRI-Values.aspx [accessed Feb 17, 2011]. [10] Poliquin S, Joseph L, Gray-Donald K. Calcium and vitamin D intakes in an adult Canadian population. Can J Diet Pract Res 2009;70:21–7. [11] Blalock SJ, Norton LL, Patel RA, Cabral K, Thomas CL. Development and assessment of a short instrument for assessing dietary intakes of calcium and vitamin D. J Am Pharm Assoc 2003;43:685–93. [12] Chee WS, Suriah AR, Zaitun Y, Chan SP, Yap SL, Chan YM. Dietary calcium intake in postmenopausal Malaysian women: comparison between the food frequency questionnaire and three-day food records. Asia Pac J Clin Nutr 2002;11:142–6. [13] Cummings SR, Block G, McHenry K, Baron RB. Evaluation of two food frequency methods of measuring dietary calcium intake. Am J Epidemiol 1987;126:796–802. [14] Haines CJ, Chung TK, Leung PC, Leung DH, Wong MY, Lam LL. Dietary calcium intake in postmenopausal Chinese women. Eur J Clin Nutr 1994;48:591–4. [15] Khan NC, Mai lB, Hien VT, Lam NT, Hoa VQ, Phuong TM, et al. Development and validation of food frequency questionnaire to assess calcium intake in postmenopausal Vietnamese women. J Nutr Sci Vitaminol (Tokyo) 2008;54: 124–9. [16] Magkos F, Yannakoulia M. Methodology of dietary assessment in athletes: concepts and pitfalls. Curr Opin Clin Nutr Metab Care 2003;6:539–49. [17] Montomoli M, Gonnelli S, Giacchi M, Mattei R, Cuda C, Rossi S, et al. Validation of a food frequency questionnaire for nutritional calcium intake assessment in Italian women. Eur J Clin Nutr 2002;56:21–30. [18] Sato Y, Tamaki J, Kitayama F, Kusaka Y, Kodera Y, Koutani A, et al. Development of a food-frequency questionnaire to measure the dietary calcium intake of adult Japanese women. Tohoku J Exp Med 2005;207:217–22. [19] Smith BA, Morgan SL, Vaughn WH, Fox L, Canfield GJ, Bartolucci AA. Comparison of a computer-based food frequency questionnaire for calcium intake with 2 other assessment tools. J Am Diet Assoc 1999;99:1579–81. [20] Uenishi K, Ishida H, Nakamura K. Development of a simple food frequency questionnaire to estimate intakes of calcium and other nutrients for the prevention and management of osteoporosis. J Nutr Sci Vitaminol (Tokyo) 2008;54:25–9. [21] BC Dairy Foundation. Calcium Calculator. Available from: http://bcdairyfoundation.ca/interactive/calcium-calculator/ [accessed Feb 17, 2011]. [22] Cheung AM, Tile L, Lee Y, Tomlinson G, Hawker G, Scher J, et al. Vitamin K supplementation in postmenopausal women with osteopenia (ECKO trial): a randomized controlled trial. PLoS Med 2008;5:e196. [23] Osteoporosis Canada. Calcium Calculator. Available from: [accessed http://www.osteoporosis.ca/index.php/ci id/5355/la id/1.htm Feb 17, 2011]. [24] Washington State Dairy Council. Calcium Calculator. Available from: http://www.eatsmart.org/games/c calculator/ [accessed Feb 17, 2011]. [25] Fred Hutchinson Cancer Research Center. Fred Hutchinson Cancer Research Center Food frequency questionnaires. Available from: www.fhcrc.org/science/shared resources/nutrition/ffq/index.html [accessed Feb 17, 2011].
172
A. Hung et al. / Maturitas 69 (2011) 168–172
[26] ESAH Research. The Food Processor. Available from: http://www.esha.com [accessed Feb 17, 2011]. [27] Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–8. [28] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10. [29] Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29–36. [30] Vatanparast H, Dolega-Cieszkowski JH, Whiting SJ. Many adult Canadians are not meeting current calcium recommendations from food and supplement intake. Appl Physiol Nutr Metab 2009;34:191–6. [31] Angbratt M, Moller M. Questionnaire about calcium intake: can we trust the answers? Osteoporos Int 1999;9:220–5. [32] Green JH, Booth CL, Bunning RL. Assessment of a rapid method for assessing adequacy of calcium intake. Asia Pac J Clin Nutr 2002;11:147–50. [33] Pasco JA, Sanders KM, Henry MJ, Nicholson GC, Seeman E, Kotowicz MA. Calcium intakes among Australian women: Geelong Osteoporosis Study. Aust N Z J Med 2000;30:21–7.
[34] Johnson-Down L, Ritter H, Starkey LJ, Gray-Donald K. Primary food sources of nutrients in the diet of Canadian adults. Can J Diet Pract Res 2006;67: 7–13. [35] Brown JL, Griebler R. Reliability of a short and long version of the block food frequency form for assessing changes in calcium intake. J Am Diet Assoc 1993;93:784–9. [36] Musgrave KO, Giambalvo L, Leclerc HL, Cook RA, Rosen CJ. Validation of a quantitative food frequency questionnaire for rapid assessment of dietary calcium intake. J Am Diet Assoc 1989;89:1484–8. [37] Sebring NG, Denkinger BI, Menzie CM, Yanoff LB, Parikh SJ, Yanovski JA. Validation of three food frequency questionnaires to assess dietary calcium intake in adults. J Am Diet Assoc 2007;107:752–9. [38] Angus RM, Sambrook PN, Pocock NA, Eisman JA. A simple method for assessing calcium intake in Caucasian women. J Am Diet Assoc 1989;89: 209–14. [39] Clover E, Miller M, Bannerman E, Magarey A. Relative validation of a short food frequency questionnaire to assess calcium intake in older adults. Aust N Z J Public Health 2007;31:450–8.