Prevalensi panjang badan lahir pendek di Indonesia masih tinggi dan menjadi masalah kesehatan masyarakat yang disebabkan oleh pelbagai faktor secara langsung dan tidak langsung serta berdampak luas dan berkelanjutan dalam siklus kehidupan. Penelitian menggunakan data Riset Kesehatan Dasar (Riskesdas) 2013 dengan pendekatan potong lintang bertujuan mengetahui mekanisme hubungan berbagai variabel laten terhadap prevalensi panjang badan lahir pendek. Sampel adalah 497 kabupaten yang diagregat dari data individu, yaitu anak lahir dari ibu berusia 15 hingga 49 tahun dengan kriteria anak kandung dan lahir tunggal. Pemodelan menggunakan Structural Equation Modelling. Kehamilan berisiko tinggi berhubungan positif langsung dengan prevalensi panjang badan lahir pendek (r = 0,279; nilai p = 0,014). Pemanfaatan pelayanan kesehatan berhubungan positif tidak langsung dengan prevalensi panjang badan lahir pendek melalui kehamilan berisiko tinggi (r = 0,135; nilai p = 0,029). Sosial ekonomi tidak berhubungan signifikan dengan prevalensi panjang badan lahir pendek (r = -0,087; nilai p = 0,156), namun akan berhubungan bila melalui mekanisme hubungan pemanfaatan pelayanan kesehatan (r = 0,653; nilai p = 0,0001) dan kehamilan berisiko tinggi (r = 0,759; nilai p = 0,0001). Upaya intervensi perlu difokuskan pada pencegahan kehamilan berisiko tinggi melalui perbaikan status gizi dan kesehatan ibu sejak usia remaja untuk menurunkan prevalensi panjang badan lahir pendek.

---

The prevalence of short birth length in Indonesia still high and it becomes a public health problem caused by any direct and indirect factors as well as having a wide and sustainable effect in life cycle. The study used Basic health system (Riskesdas) 2013 data with a cross-sectional approach aiming to find out the mechanism of the relation between any latent variables to the short birth length prevalence. Samples were 497 districts aggregated from individual data that were children children born by 15 - 49 year-old mothers with biological children and single birth criteria. The modelling used Structural Equation Modeling. High-risk pregnancy had a direct positive relation with the prevalence of short birth length (r = 0.279; p value= 0.014). The use of health services had an indirect positive relation with short birth length prevalence through high-risk pregnancy (r = 0.135; p value= 0.029). Social economy did not have any significant relation with the prevalence of short birth length (r = -0.087; p value = 0.156), but would be related if through the mechanism of health service use (r = 0.653 ; p value = 0.0001) and high-risk pregnancy (r = 0.759 ; p value = 0.0001). Efforts of intervention need to be focused on prevention of high-risk pregnancy through improvement of nutritional and health status of mothers since teenager in order to reduce short birth length prevalence.