Development, validation, and use of a semi-quantitative food frequency questionnaire for assessing protein intake in Papua New Guinean Highlanders

• Published in: American journal of human biology Vol. 27; no. 3; pp. 349 - 357
• Main Authors: Morita, Ayako, Natsuhara, Kazumi, Tomitsuka, Eriko, Odani, Shingo, Baba, Jun, Tadokoro, Kiyoshi, Igai, Katsura, Greenhill, Andrew R., Horwood, Paul F., Soli, Kevin W., Phuanukoonnon, Suparat, Siba, Peter M., Umezaki, Masahiro
• Format: Journal Article
• Language: English
• Published: United States Blackwell Publishing Ltd 01.05.2015; Wiley Subscription Services, Inc
• Subjects: Adolescent; Adult; Age Factors; Aged; Biomarkers; Body Weights and Measures; Child; Diet Surveys - methods; Diet Surveys - standards; Dietary Proteins - analysis; Energy Intake; Ethnic Groups; Female; Hair - chemistry; Humans; Male; Middle Aged; Papua New Guinea; Reproducibility of Results; Sex Factors; Socioeconomic Factors; Papua New Guinea
• ISSN: 1042-0533; 1520-6300
• DOI: 10.1002/ajhb.22647

Objectives The aim of this article was to develop a semi‐quantitative food frequency questionnaire (FFQ) and evaluate its validity to estimate habitual protein intake, and investigate current dietary protein intakes of Papua New Guinea (PNG) Highlanders. Methods A 32‐item FFQ was developed and tested among 135 healthy male and female volunteers. The FFQ‐estimated daily total and animal protein intakes were compared with biomarkers and 3‐day Weighed Food Records (WFR) by correlation analyses, Bland–Altman plot analyses and joint classification analyses. Results The FFQ‐estimated total protein intake significantly correlated with urinary nitrogen in the first morning void after adjusting urinary creatinine concentration (r = 0.28, P < 0.01) and the FFQ‐estimated animal protein intake significantly correlated with the hair δ15N (Spearman's r = 0.34, P < 0.001). The limits of agreement were ±2.39 Z‐score residuals for total protein intake and ±2.19 Z‐score for animal protein intake, and intra‐individual differences increased as protein intake increased. The classification into the same and adjacent quartiles was 66.0% for total protein intake and 73.6% for animal protein intake. Median daily total and animal protein intake estimates from the FFQ and the 3‐day WFR showed a good agreement with differences of 0.2 and 4.9 g, respectively. None of the studied communities in the PNG Highlands met the biologically required protein intake; although the community closer to an urban center showed higher protein intake than the more remote communities. Conclusions The newly developed 32‐item FFQ for PNG Highlanders is applicable for evaluation of protein intake at the individual level. Am. J. Hum. Biol. 27:349–357, 2015. © 2014 Wiley Periodicals, Inc.