Indian Journal of Community Medicine (Aug 2024)

Cobalamin Intake in North Indians by Food Frequency Questionnaire (COIN-FFQ) – A Development and Validation Study

  • Swapnil Rawat,
  • Meena Kumari,
  • Jitender Nagpal

DOI
https://doi.org/10.4103/ijcm.ijcm_155_23
Journal volume & issue
Vol. 49, no. 5
pp. 726 – 733

Abstract

Read online

Background: Vitamin B12 deficiency is widely prevalent in all age groups which is of major concern. However, there is no valid Food Frequency Questionnaire (FFQ) for dietary vitamin B12 estimation. Hence, we aimed to develop and validate an FFQ for the estimation of dietary intake of Vitamin B12. Materials and Methods: Commonly consumed B12-rich food items were selected from literature and filtered using a market survey. For concordant and discriminant validation, B12 and homocysteine levels were estimated. To establish convergent validity, the Cobalamin Intake in North Indians by Food Frequency Questionnaire (COIN-FFQ) and 72-hour dietary recall (72HrDR) were both administered to the same subjects. The COIN-FFQ was readministered after initial administration for test–retest reliability. Internal consistency of the FFQ was then tested using Cronbach’s alpha. Results: We enrolled 115 adults with a mean age and weight of 31.9 ± 8.7 years and 66.0 ± 11.8 kg, respectively. In total, 19.1% were vegetarian. The dietary B12 using COIN-FFQ (n = 60; mean = 4.3 ± 1.8 µg/d) was significantly correlated (r = 0.255; P = 0.049) with serum levels (mean = 120.1 ± 62.6 pmol/L) establishing concordant validity. A significant difference was noted between the dietary, serum B12, and homocysteine levels of vegetarians versus nonvegetarians establishing discriminant validity (mean diff 1.4 (0.5–2.4), P = 0.004; Z-statistic −2.182, P value 0.029, and Z-statistic −2.438; P value 0.015), respectively. FFQ was strongly correlated with 72HrDR and test–retest FFQ (n = 27; r = 0.814, P < 0.001 and r = 0.869, P < 0.001, respectively) establishing convergent validity and test–retest reliability. The internal consistency with Cronbach’s alpha was in the acceptable range, 0.631 (n = 115). Conclusion: The newly developed COIN-FFQ is valid and reliable in estimating dietary B12 intake.

Keywords