Open Access Research article

A prospective cohort study of stroke mortality and arsenic in drinking water in Bangladeshi adults

Mahfuzar Rahman12*, Nazmul Sohel3, Mohammad Yunus1, Mahbub Elahi Chowdhury1, Samar Kumar Hore1, Khalequ Zaman1, Abbas Bhuiya1 and Peter Kim Streatfield1

Author Affiliations

1 ICDDRB, 68 Shahid Tajuddin Ahmed Sarani, Mohakhali, Dhaka 1212, Bangladesh

2 University of Chicago Research Bangladesh, House 338, Road 24, New DOHS, Mohakhali, Dhaka 1212, Bangladesh

3 Department of Clinical Epidemiology and Biostatistics, McMaster University, 1280 Main Street West, Hamilton ON L8S 4K1, Ontario, Canada

For all author emails, please log on.

BMC Public Health 2014, 14:174  doi:10.1186/1471-2458-14-174

Published: 18 February 2014



Arsenic in drinking water causes increased coronary artery disease (CAD) and death from CAD, but its association with stroke is not known.


Prospective cohort study with arsenic exposure measured in well water at baseline. 61074 men and women aged 18 years or older on January 2003 were enrolled in 2003. The cohort was actively followed for an average of 7 years (421,754 person-years) through December 2010. Based on arsenic concentration the population was categorized in three groups and stroke mortality HR was compared to the referent. The risk of stroke mortality Hazard Ratio (HR) and 95% Confidence Interval was calculated in relation to arsenic exposure was estimated by Cox proportional hazard models with adjustment for potential confounders.


A total of 1033 people died from stroke during the follow-up period, accounting for 23% of the total deaths. Multivariable adjusted HRs (95% confidence interval) for stroke for well water arsenic concentrations <10, 10-49, and ≥50 μg/L were 1.0 (reference), 1.20 (0.92 to 1.57), and 1.35 (1.04 to 1.75) respectively (Ptrend=0.00058). For men, multivariable adjusted HRs (95%) for well water arsenic concentrations <10, 10-49, and ≥50 μg/L were 1.0 (reference), 1.12 (0.78 to 1.60), and 1.07 (0.75 to 1.51) respectively (Ptrend=0.45) and for women 1.0 (reference),1.31 (0.87 to 1.98), and 1.72 (1.15 to 2.57) respectively (Ptrend=0.00004).


The result suggests that arsenic exposure was associated with increased stroke mortality risk in this population, and was more significant in women compared to men.