Reliability and Validity of a Brief Questionnaire to Assess Calcium Intake of Middle-School-Aged Children

Division of Epidemiology and Community Health, School of Public Health, University of Minnesota, Minneapolis 55454, USA.
Journal of the American Dietetic Association (Impact Factor: 3.92). 11/2006; 106(11):1790-5. DOI: 10.1016/j.jada.2006.08.014
Source: PubMed


Evaluate the validity and reliability of a short (10-item) calcium food frequency questionnaire (FFQ) for use with middle-school-aged (11 to 14 years of age) children.
The calcium FFQ was completed twice, with 1 week between administrations. Three 24-hour dietary recalls were collected from each participant after the second administration of the calcium FFQ.
Students in an ethnically diverse middle school in Minneapolis, MN (n=248).
Calcium intake estimates from the calcium FFQ and dietary recalls.
Correlations between calcium intake estimates from the first and second questionnaire administrations of the calcium FFQ were calculated and paired t tests were conducted to compare mean calcium intake estimates from each questionnaire administration. Mean intake estimates from the calcium FFQ and the dietary recalls were compared. Also, correlations between intake estimates from the calcium FFQ and the recalls were calculated.
Correlation between calcium intake estimates derived from the first and second administration of the calcium FFQ was 0.74. Mean calcium intake estimates from the calcium FFQ and the average of the three dietary recalls were 856 mg/day and 993 mg/day, respectively (P<0.001). The correlation between calcium intake estimates derived from the calcium FFQ and the average of the recalls was 0.43.
Reliability of the FFQ was found to be good while validity was weaker, with calcium intake from the calcium FFQ moderately associated with estimates from dietary recalls. Where a brief instrument for assessing calcium intake of middle-school-aged children is needed, the calcium FFQ evaluated in this study may be useful.

13 Reads
  • Source
    • "Other studies have included either no calcium-fortified foods (Magkos et al, 2006; Cook et al, 2003; Yanek et al, 2001; Jensen et al, 2004; Blalock et al, 1998), calciumfortified mineral water (Montomoli et al, 2002), calcium-fortified juice alone (Harnack et al, 2006), calcium-fortified juice and a grain product (Ward et al, 2004), or did not provide details regarding the inclusion of calcium-fortified foods in the assessment tool (Sebring et al, 2007; Bell et al, 2002; Blalock et al, 2003; Musgrave et al, 1989; Chee et al, 2002; Taitano et al, 1995; Xu et al, 2000). In contrast, this CFFFQ includes 14 fortified food choices including 1 in fruits, 8 in grains, and 5 in other foods. "
    [Show abstract] [Hide abstract]
    ABSTRACT: The purpose of this research was to measure and evaluate lifestyle changes affecting bone health. Osteoporosis is a public health concern. Improving bone health, thereby preventing osteoporosis and fractures, can lead to a reduction in health care costs. Calcium is a key player in bone health. With an increase in calcium-fortified foods an objective was to develop a food frequency questionnaire that assesses both natural and fortified sources of calcium. A validation study in which the Calcium-Focused Food Frequency Questionnaire (CFFFQ) was tested against a 24-hour recall in adult females (pilot study, n=15) and college-aged females (primary study, n=300). In the pilot study, no significant differences in calcium intake for total calcium or food group category was found except for calcium from “foods with dairy” (t=2.23, p=.043) and “vegetable” (t=-3.106, p=.008). In the primary study and after removal of outliers (n= 187), significant correlations (r=.155 to.74, p<.04) were found between calcium (mg) in CFFFQ and 24-hour recall for “dairy”, “foods with dairy”, “fruit”, “vegetables”, “grains” groups and total calcium. In the reliability study, all groups were significantly correlated (r=.155 to .96, p<.034) except for the dairy. In using the CFFFQ with post-menopausal women [46 black and 139 white post-menopausal women (age 69.4 +5.8 years)], as daily calcium intake increased, the 24-hour recall increasingly underreported calcium (r = .41, p<.001). Per cross-tabulation and Chi-square analyses, the CFFFQ had greater specificity for lower calcium intakes. For calcium classified by food groups, there was moderate correlation for dairy (r = .56, p<.001) and fruit groups (r = .434, p<.001). Dairy was the primary calcium source for both groups (55% and 57% of intake for black, white women, respectively). The CFFFQ can be used to identify those with inadequate calcium intakes (<800 mg/day) and to identify key sources of dietary calcium. The CFFFQ was used in part of the larger bone-health community program (8 weeks) addressing disease risk and lifestyle changes within the framework of behavior constructs (n=69). There was significant increases in calcium intake (p<.027) and vitamin D intake (p<.015), with calcium from the fruit group (p<.005, 24-hour recall) and grain group (p<.042, CFFFQ). There was a significant change (p<.01) in 3 of 5 items related to susceptibility; 3 of 3 items related to perceived severity (p<.03); in 5 of 5 items related to benefits of nutrition changes (p<.001); in 1 of 7 items related to nutrition barriers (p<.05); in 4 of 4 nutrition self-efficacy items (p<.01); in 4 of 6 items related to subjective norm (p<.05); in 4 of 5 nutrition attitudes (p<.05) and 3 of 4 intentions (p<.01) [Wilcoxon Signed Rank]. This theory-based program was successful in improving calcium intake, vitamin D intake and Health Belief Model (HBM) and Theory of Reasoned Action (TRA) constructs related to bone-healthy diets, implying effective program applications to clinic and community-based practice.
    Preview · Article · Jan 2010
  • Source
    • "Use of FFQ for assessing children's diets possesses extra challenges beyond those for assessing diets of adults, because children may have more limited cognitive ability and knowledge of food and food preparation. Development and evaluation of FFQ for children has been continuously conducted for various areas of diets in an effort to provide a useful dietary assessment tool during the last decade (Buzzard et al., 2001; Cullen et al., 1999; Field et al., 1999; Harnack et al., 2006; Jensen et al., 2004; Rockett et al., 1995; Vereecken & Maes, 2003; Watson et al., 2003). However, few systematic studies have been nationally done to develop and test a FFQ for Korean school-aged children. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This study sought to compare intake levels of major dish items and food groups between normal and obese Korean children in order to understand obesity-related eating behaviors. The study was also conducted to apply the information on the differences in major dish items and food groups found in this study to development of a brief food frequency questionnaire (FFQ) for children's obesity-related eating behaviors. Finally, the developed FFQ was evaluated for reliability. A total of 290, 24-hour dietary recall data from 9-11 year-old normal or obese participants in the 2001 Korean National Health and Nutrition Examination Survey (KNHANES) were reviewed to identify dish items and food groups showing outstanding differences between the normal and obese groups. Based on the level of intake amount between the two groups, a total of 7 items, including ddeokbokggi, spaghetti, ham roast, pork cutlet, dairy products & ices, kimchi, and fruits, were selected to be included in the FFQ. The former 5 items were seen to be consumed more in the obese group, while the latter 2 items were so in the normal group. The questionnaire was formatted into a frequency response section of a seven-category option and reference period of the last 7 days. Test-retest reliability of the developed FFQ was examined by administering it to 153, 9-11 year-old children at a public elementary school in Seoul twice at a month interval. The level of reliability was found to be reasonably high. In conclusion, this study suggests that high consumption of several high-calorie dish items and low consumption of kimchi and fruits may be important eating behaviors in relation to obesity risk among 9-11 year-old children who need nutrition educators' attention. The current study also implies that a simple FFQ can be utilized to reliably assess 9-11 year-old children's diets.
    Full-text · Article · Dec 2007 · Nutrition research and practice
  • Source
    • "A number of recent studies have also used 24-hour recalls to validate different calcium FFQs [20-22]. In accordance with our results, Harnack et al. [21] also reported that although the test-retest reliability of their 10-item calcium FFQ was good (ICC = 0.66), the calcium intake values from FFQ was only moderately associated with estimates from 24-hour recalls (ICC = 0.40) in 11–14 years old boys. Magkos et al. [22] found that their 30-item calcium FFQ significantly underestimated the calcium intake in 351 Greek children aged 11.9 ± 1.2 years. "
    [Show abstract] [Hide abstract]
    ABSTRACT: In order to rapidly assess nutrient intake, Food Frequency Questionnaires (FFQ) have been developed and proven to be reliable for quick, user friendly analysis in adults. However, the accuracy of these questionnaires in children has been studied to a limited extent. The aim of this study was to compare the daily calcium intake values obtained from the Rapid Assessment Method (RAM), an FFQ, for assessing daily calcium intake in child and adolescent males with the values obtained from the 24-hour recall method. Subjects included 162 child and adolescent males, aged 9-16 years, subdivided into elementary school (ES, 9-12 years) and high school (HS, 14-16 years) age groups. Daily calcium intake was significantly lower in ES compared with HS, using both methods. The intra-class correlation coefficients (ICC) between RAM values and those obtained using the 24-hour recall questionnaire were significant yet moderate (ICC = 0.46 and 0.43 for ES and HS, respectively). However, daily calcium intake obtained using RAM was significantly higher when compared with the 24-hour recall values in both ES (1576 +/- 1101 vs. 1003 +/- 543 mg, in RAM and 24-hour, respectively) and in HS males (1873 +/- 739 vs. 1159 +/- 515 mg, in RAM and 24-hour, respectively). RAM overestimates daily calcium intake as compared with the 24-hour recall method in both child and adolescent males.
    Full-text · Article · Feb 2007 · Nutrition Journal
Show more