Article Text


Chronic disease
P2-113 Evaluation of vitamin D deficiency determinants in urban areas of Iran by generalised estimating equations analysis method
  1. R Heshmat1,
  2. K Mohammad2,
  3. S R Majdzadeh2,
  4. B Larijani1,
  5. M H Forouzanfar2
  1. 1Endocrinology & Metabolism Research Institute, Tehran University of Medical Sciences, Tehran, Iran
  2. 2School of Public Health, Tehran University of Medical Sciences, Tehran, Iran


Introduction Vitamin D plays an integral role in bone mineralisation. Its deficiency has been shown to be associated with some cancers, cardiovascular disease, diabetes and osteoporosis. We aimed to evaluate the factors determining vitamin D levels using Generalised Estimating Equation (GEE). Its main application is evaluation of related data in longitudinal and hierarchal states, especially in cluster samples which can result in an unbiased estimation.

Methods In a random cluster sample, 5232 subjects from five urban areas (Tehran, Tabriz, Mashhad, Shiraz and Booshehr) were recruited. A fasting blood sample was taken for measurement of 25-hydroxy vitamin D levels.

Results In the GEE model, age group, sun block usage, use of Islamic coverage and geographical variables were removed from the model as was city of residence (as it was collinear with geographical and environmental factors), sex and the interaction of age and sex group were correlated with vitamin D deficiency. Living in Tehran, Mashhad and Shiraz was associated with vitamin D deficiency [OR (95% CI) 2.1 (1.7 to 2.5), 0.9 (0.7 to 1.1) and 0.7 (0.5 to 0.9) respectively]. The ratio for males to females was 1.3 (1.1 to 1.6).

Conclusion Analyses showed that environmental factors in residential locations, female sex and the interaction of sex and age are protective on vitamin D levels. Moreover, analysis by GEE method compared to logistic regression did not show any significant variation in the results which indicate that variation in vitamin D deficiency is due to differences between factors such as location and sex rather than deviation in samples of each cluster.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.