Poster Topical Area: Nutritional Epidemiology

Location: Hall D

Poster Board Number: 842

P20-097 - Association between estimated usual intake of fish and seafood and osteoporosis: application of the statistical model in estimating the usual intake from two 24-hour recalls

Monday, Jun 11
8:00 AM – 3:00 PM

Objectives: We examined the associations between estimated usual intakes of fish and seafood and the prevalence of osteoporosis by adjusting the effect of within-person variation in intake.


Methods: We included a total of 6639 adults from Korean National Health and Nutrition Examination Survey (KNHANES) 2008-2011. We adjusted the effect of within-person variation in fish and seafood intake using data from 2-day 24 hour recalls of KNHANES 2009 by applying statistical analysis modified from the model developed by National Institute of Food and Drug Safety Evaluation. Bone mineral density (BMD) of the lumbar spine, hip, or femoral neck was measured using dual-energy X-ray absorptiometry and osteoporosis was defined as BMD T-scores less than -2.50. We used the multivariable logistic regression models and calculated odds ratio (OR)s and 95% confidence interval (CI)s to examine the associations between intakes of fish and seafood and the prevalence of osteoporosis.


Results: We found that the associations of estimated usual intakes of fish and seafood with the prevalence of osteoporosis were more pronounced than those of intakes from 1-day 24 hour recalls. When we estimated fish intake from 1-day 24 hour recalls, ORs (95% CI)s were 0.80[0.61-1.04] for 15-<30 g/d, 0.77[0.55-1.09] for 30-<45 g/d, and 0.92[0.76-1.13] for ≥45 g/d compared with <15g/d of fish intake. However, estimation of the distributions of usual fish intakes de-attenuated the associations; compared with <15 g/d of estimated usual intake of fish, OR (95% CI)s were 0.78 [0.64-0.95], 0.72[0.57-0.91], and 0.64[0.49-0.84] for the same categories, respectively (p for trend < 0.01). Similarly, we found a significant association for estimated usual intakes of seafood, but did not find a significant trend for seafood intake from 1-day 24 hour recalls; compared with <30 g/d, OR (95% CI)s were 0.94[0.79-1.11], 0.81[0.65-1.00], 0.76[0.56-1.02] for 30-<50 g/d, 50-<80 g/d, and ≥80 g/d of estimated usual seafood intakes, respectively (p for trend < 0.05).


Conclusions: We found that usual intakes of fish and seafood were inversely associated with the prevalence of osteoporosis in Korean adults.




Funding Source:

Acknowledgements: This research was supported by a grant (17162MFDS026) from Ministry of Food and Drug Safety in 2018.

CoAuthors: Jiyoung Youn – Seoul National University, Seoul, Korea; Sang Ah Chi – Sungkyunkwan University, Seoul, South Korea; Jeong-sun Ahn – Seoul National University, Seoul, Korea; Sang-eun Moon – Seoul National University, Seoul, Korea; In-Kwon Yeo – Sookmyung Women’s University, Seoul, Korea; Dong Woo Kim – Korea National Open University, Seoul, Korea; Kyunga Kim – Research Institute for Future Medicine, Samsung Medical Center, Seoul, South Korea; Jung Eun Lee – Seoul National University, Seoul

Heejin Lee

Graduate student
Seoul National University
Daejeon, Taejon-jikhalsi, Republic of Korea