Study type: Epidemiological study (observational study)

Cause-specific mortality in cellular telephone users. epidem.

Published in: JAMA 1999; 282 (19): 1814-1816

Aim of study (acc. to author)

Cause-specific mortality in cellular telephone users was investigated in an expanded cohort study (see publication 1905).

Further details

Two different types of cellular phones were included in the study: the handheld model which has the antenna in the hand piece in close proximity to the head during a call, and the nonhandheld model, mainly car telephones. Users of nonhandheld model are considered as unexposed because the antenna is not part of the hand set.

Endpoint/type of risk estimation

Type of risk estimation: (standardized mortality rate (SMR))



Exposure groups

Group Description
Reference group 1 nonhandheld telephones, calling time: median 1.5 min/day
Group 2 handheld telephones, calling time: < 2 min/day, median 0.8 min/day
Group 3 handheld telephones, calling time: ≥ 2 min/day, median 5.0 min/day
Reference group 4 nonhandheld telephones, length of service: median 2.1 years
Group 5 handheld telephones, length of service: ≤ 3 years, median 1.6 years
Group 6 handheld telephones, length of service: > 3 years, median 3.8 years


Study size

Type Value
Total 285,561

Conclusion (acc. to author)

The only category of cause of death for which there was indication of increasing risk with increasing minutes of cellular telephone use was "motor vehicle collisions". Similar results were found for number of calls per day. The distinction between handheld and nonhandheld telephones did not apply since both telephone types might interfere with driving.
To investigate the relation between cellular telephone use and risk of fatal motor vehicle collisions, the authors combined handheld and nonhandheld cellular telephone users and looked for trends. They found increasing risk of mortality with increasing telephone use, but an inverse trend with number of years of service.

Limitations (acc. to author)

Because the analyses were based on billing data, some information were missing: which calls were placed from a motor vehicle, whether a cellular telephone was used immediately preceding the fatal motor vehicle crash, how many hours individuals drove and the relation between driving hours and telephone hours.

Study funded by

Related articles