## Table 5 |
|||

Parameter estimates and standard errors from linear mixed effects model predicting
linear growth (height increase) of boys over the follow up period |
|||

Effect (n=1431) |
Estimate |
SE |
P |

Fixed Effects |
|||

Food Insecurity | −0.3382 | 0.3405 | 0.3210 |

Height for age z-scores at baseline | 6.6557 | 0.0911 | <.0001 |

Food Insecurity X Time | 0.1420 | 0.2904 | 0.6249 |

Time | 1.3737 | 0.4000 | 0.0006 |

Age of boys at baseline | 3.9514 | 0.0804 | <.0001 |

Household Income Tertile | |||

High (Reference) | |||

Middle | −0.3571 | 0.3695 | 0.3343 |

Middle X time | 0.6131 | 0.3522 | 0.0822 |

Low | 0.1416 | 0.3864 | 0.7140 |

low X time | −0.1378 | 0.3663 | 0.7069 |

Place of Residence | |||

Urban (Reference) | |||

Semi-urban | −0.3955 | 0.3546 | 0.2651 |

Semi-urban X time | 0.9695 | 0.3396 | 0.0040 |

Rural | −0.6366 | 0.3717 | 0.0873 |

Rural X Time | 1.1587 | 0.3546 | 0.0011 |

Quadratic time effect (Time^{2}) |
1.9663 | 0.1565 | <.0001 |

Random effects |
|||

Variance | 14.6274 | 1.5378 | |

CS | −5.1433 | 0.8679 | |

Variance of measurement errors (residuals) | 11.4764 | 0.4959 |

Restricted maximum likelihood (REML) method was used in estimating the parameters.

Compound symmetry (CS) variance matrix structure was used to avoid over parameterization of the estimates.

SE= Standard error.

Time = follow up rounds.

Belachew * et al.*

Belachew * et al.* *Nutrition Journal* 2013 **12**:55 doi:10.1186/1475-2891-12-55