Introduction Osteoporosis has become a widespread disease. 1 Weakening bones cause about 8.9 million fractures per year worldwide; around 200 million women are osteoporotic patients. 2 In Pakistan, this 'silent disease' affects 6.7 million people and it will increase to more than 7.1 million by 2020. 3 Majority of cases are not detected due to unavailability of dual-energy X-ray absorptiometry (DEXA) machines. Only 16 DEXA scanners are available in Pakistan, 4 rendering the cost of scans beyond the reach of most people in Karachi. According to the World Health Organisation (WHO), osteoporosis is referred to as faulty and weakened bone structure due to low bone mineral content per unit volume. 5 Reduction of bone mass decreases strength and quality of bone, resulting in an increased fracture risk. The WHO declared DEXA scan as the gold standard to diagnose weakening bones. 5 The bone mineral density (BMD) values are presented asT-scores and Z-scores; these being the number of standard deviations above or below an age-matched population (Z-scores) and above or below a young healthy adult population of the same ethnicity (T- scores). 6 WHO defined osteoporosis as T-score less than - 2.5 of lumbar spine, hip or distal forearm, 7 but the diagnostic meaning of BMD values by DEXA scanner is different between different DEXA manufacturers. 8 Further, the subcontinent, especially Pakistan, still faces the same problem, as no Pakistani reference data exists for interpretation of osteoporosis. The American and the European regions resolve these problems to some extent by studying different population groups. The study based on the secondary analysis of already published data of NHANES 9 that contains the BMD data of total body, lumbar spine and thoracic spine to establish the behaviour of BMD in different age group, sex and races. This present analysis is being done, in order to compare a future local study with similar variables. Subjects and Methods The BMD of total 26,215 subjects were measured by DEXA. The survey was conducted in America by National Health and Nutrition Examination Survey (NHANES) from 1999-2006. 9 The subjects were categorized into gender, Vol. 65, No. 4, April 2015 388 ORIGINAL ARTICLE Bone mineral density comparison of total body, lumbar and thoracic: an exploratory study Sheikh Kashif Raffat, Abdul Basit Shaikh, Muhammad Sarim, Akhter Raza Syed Abstract Objective: To analyse the behaviour of bone mineral density (BMD) of total body, lumbar and thoracic spine, and compare it between different age-groups, ethnicity and gender as a secondary analysis of already published data of NHANES. Methods: The study was done and compared data of the National Health and Nutrition Examination Survey which was done from 1999 to 2006. It comprised bone mineral density data of 26,215 subjects of age 8 years and over. The sample was divided into 10 age groups, 3 ethnicities and gender groups. All subjects had bodyweights not exceeding 300lb, and height not greater than 6'5''. Female subjects were non-pregnant. SPSS 16 was used for statistical analysis. Result: Of the total subjects, 7712 (29.41%) were Mexican Americans, 11236 (42.86%) were non-Hispanic whites and 7267 (27.73%) were non-Hispanic blacks. There were 13291 (50.69%) males, and each ethnic group also had approximately 50% males. The bone mineral density of non-Hispanic blacks was higher than the other two groups, whereas males of all ethnic groups and races had higher density then females of the same group. Bone mineral density of total body, lumbar and thoracic was significantly different. After 50 years of age, total body, lumbar and thoracic bone mineral density decreased in females but lumbar slightly increased in males, whereas thoracic remained stable and total body decreased. Conclusion: The bone mineral density of male and blacks was higher than female and non-black races in total body, lumbar and thoracic spine. The thoracic BMD was lower and might predict early fracture risk. Keywords: Bone mineral density, DEXA, Lumbar spine, Osteoporosis, Thoracic spine. (JPMA 65: 388; 2015) Department of Computer Science, Federal Urdu University of Arts, Sciences and Technology, Karachi, Pakistan. Correspondence: Sheikh Kashif Raffat. Email: kashifraffat@fuuast.edu.pk