[ABSTRAK
Informasi mengenai adanya semi-quantitative food frequency questionnaire(SFFQ) yang valid untuk mengukur asupan PUFA pada anak di Indonesia masihminim. Oleh karena itu, tujuan dari penelitian ini adalah untuk mengembangkanSFFQ dan menguji validitas dan reliabilitas dari SFFQ tersebut untuk mengukurasupan PUFA pada anak di Indonesia usia 6-23 bulan.Penelitian ini dilaksanakan dengan desain cross sectional di dua kelurahanwilayah Jakarta Timur. Penelitian ini melibatkan 89 anak yang dipilih secaramultistage random sampling. Anak-anak tersebut dilakukan pengukuran intakemelalui SFFQ dan wawancara recall selama 3 hari tidak berturut-turut. Beberapaanak (n=35) telah dipilih untuk dilakukan pengukuran plasma lipid dalam darah.Formulir SFFQ terdiri dari 78 item makanan yang disusun dari tabel komposisibahan makanan luar Indonesia. SFFQ tersebut kemudian divalidasi denganwawancara recall dan plasma lipid dalam darah. Pelaksanaan SFFQ dilakukan duakali dengan rentang 4 minggu terpisah untuk mengetahui reliabilitasnya. Relatifvaliditas dan realibilitas dari SFFQ disimpulkan dari hasil analisa Bland-Altman.Uji korelasi parsial yang telah dikontrol dengan status gizi dan usia anakdilakukan untuk mengukur absolut validitas dari SFFQ.Kecocokan yang baik ditemukan antara hasil SFFQ dan wawancara recall untukDHA, EPA, DPA dan AA akan tetapi tidak untuk total n-3, n-6, ALA dan LA.Lebih jauh, SFFQ menunjukkan korelasi yang moderat dengan plasma lipid dalamdarah untuk n-6 dan LA (r: 0.40; p=0.025 and r:0.42 p=0.018, secara berurutan).Hasil analisa Bland-Altman menunjukkan 95% kecocokan antara hasil SFFQpertama dan pengulangan SFFQ untuk semua asam lemak esensial. Secarakeseluruhan, SFFQ yang dikembangkan relatif valid untuk mengukur asupanPUFA kecuali untuk total n-3, n-6, ALA dan LA. SFFQ juga reliable untukmengetahui asupan PUFA pada anak.

---

ABSTRACT
The information on the existing validated semi-quantitative food frequencyquestionnaire (SFFQ) to capture the intakes of PUFA for Indonesian children islacking. Therefore, this study aimed to investigate the validity and reproducibilityof developed SFFQ for assessing poly-unsaturated fatty acids (PUFA) intake forIndonesian children aged 6-23 months.A cross sectional study was conducted in two sub-districts of East Jakartainvolving 89 healthy children selected by multistage random sampling. Thesechildren were assessed by SFFQ and 3-day non consecutive 24-h recall. Somechildren (n=35) were randomly selected for plasma assessment (PA). The SFFQconsist of 78 food items which were constructed from the non-Indonesian foodcomposition database. It was validated against dietary recall and PA. Repeatedadministration of SFFQ (4-week apart) was conducted to assess thereproducibility of SFFQ. The relative validity and reproducibility of SFFQ weredetermined by Bland-Altman analysis. The adjusted correlation for childrennutritional status and age was performed to assess absolute validity of SFFQ.Good agreement was found between SFFQ and dietary recall for DHA, EPA,DPA, and AA, but not for total n-3, n-6, ALA and LA. Moreover, SFFQ showedmoderate correlations with plasma n-6 LCPUFA and LA (r: 0.40; p=0.025 andr:0.42 p=0.018, respectively). A 95% level of Bland-Altman agreement wasclearly observed between first SFFQ and repeated SFFQ for all essential fattyacids. In conclusion, the SFFQ was relatively valid to assess usual PUFA intakeexcept for total n-3, n-6, ALA and LA and reproducible to estimate PUFA intakeof children, The information on the existing validated semi-quantitative food frequencyquestionnaire (SFFQ) to capture the intakes of PUFA for Indonesian children islacking. Therefore, this study aimed to investigate the validity and reproducibilityof developed SFFQ for assessing poly-unsaturated fatty acids (PUFA) intake forIndonesian children aged 6-23 months.A cross sectional study was conducted in two sub-districts of East Jakartainvolving 89 healthy children selected by multistage random sampling. Thesechildren were assessed by SFFQ and 3-day non consecutive 24-h recall. Somechildren (n=35) were randomly selected for plasma assessment (PA). The SFFQconsist of 78 food items which were constructed from the non-Indonesian foodcomposition database. It was validated against dietary recall and PA. Repeatedadministration of SFFQ (4-week apart) was conducted to assess thereproducibility of SFFQ. The relative validity and reproducibility of SFFQ weredetermined by Bland-Altman analysis. The adjusted correlation for childrennutritional status and age was performed to assess absolute validity of SFFQ.Good agreement was found between SFFQ and dietary recall for DHA, EPA,DPA, and AA, but not for total n-3, n-6, ALA and LA. Moreover, SFFQ showedmoderate correlations with plasma n-6 LCPUFA and LA (r: 0.40; p=0.025 andr:0.42 p=0.018, respectively). A 95% level of Bland-Altman agreement wasclearly observed between first SFFQ and repeated SFFQ for all essential fattyacids. In conclusion, the SFFQ was relatively valid to assess usual PUFA intakeexcept for total n-3, n-6, ALA and LA and reproducible to estimate PUFA intakeof children]