Section 9: Metabolic Diseases

9.1 Diabetes Mellitus

Diabetes Mellitus and Driving Literature Review

The following section provides a review of the studies examining the influence of diabetes mellitus on driving behavior. The studies can be roughly grouped into two time periods: the early studies, which were conducted in the mid to late 60s, and more recent studies, published between 1988 and 1991. A summary of the studies is provided in Table 24.

As can be seen, some studies fail to find a significant increase in crash risk for individuals with diabetes mellitus, others report a significant increase, whereas others show no difference. Reasons for the discrepant results are proposed below.

One of the first studies to examine the risks associated with driving a motor vehicle by individuals with diabetes was conducted by Waller in 1965. In a retrospective study, he compared the driving records of 257 individuals known to the California Department of Motor Vehicles with the records of 922 randomly selected controls. For both samples, information was obtained, through direct interviews or written questionnaires, regarding age, sex, marital status, occupation, and number of miles driven annually. Results revealed that drivers with medical conditions had significantly higher crash and traffic violation rates at all ages than did those in the comparison sample. In those individuals with diabetes (type not specified), there was a reported 78 percent increase in crash rates (8.7/106 miles versus 15.5/106 miles), and a 39 percent increase in traffic violations (3.3/105 miles versus 4.6/106 miles). In Sweden, Ysander (1966) examined the driving records of individuals with chronic diseases during a ten-year period. Frequency of crashes and serious traffic offenses was compared with a control group matched for sex, age, and duration-of-license holding. In contrast to Waller's results, Ysander's results revealed a lower rate of crash and violation involvement for drivers with chronic medical conditions compared to controls. Reported crashes for individuals with IDDM were 5 percent compared to 7.7 percent for the whole control series. The rates for reported serious driving offences were 12 percent for IDDM individuals compared to 15.3 percent for controls. It is interesting to note that the average number of kilometers driven per year was substantially less for individuals with chronic diseases 26 years and older compared to their same age counter-parts. The medically impaired group aged 26-50 drove are ported 3,600 fewer kilometers per year and the 50 years of age and older group drove 13,300 fewer reported kilometers per year than their same aged peers. In addition, 21 percent of those in the medically impaired group did not drive during the study period due to illness, lowering the risk for this portion of the group. Therefore, exposure may have been a factor in the reduced rate of crashes and violations for the medically impaired group in this investigation.

Table 24 Summary of Studies on the Risk of Crash for Drivers with Diabetes Mellitus

Using the same methodology, Crancer and McMurray (1968) compared the crash and violation rates of medically restricted drivers to 1.6 million non-medically restricted licensed drivers. Individuals with diabetes (type not specified) had statistically higher crash rates (31.45 per 100 drivers), compared to age- and sex-matched controls (26.5 per 100 drivers). Amount of driving exposure was not measured in this investigation. Finally, Davis, Wahling, and Carpenter (1973) examined the driving records of individuals who had been granted driver's licenses following a review by the Oklahoma Medical Advisory Committee in 1969. The crash rates of 108 diabetic drivers the year following the review were compared to 1.65 million age- and sex-matched controls. There were no significant differences between the two groups in incidence of crashes per 100 drivers (7.4 crashes per 100 diabetic drivers versus 7.1 for the control group).

Fifteen years later, Songer, LaPorte, Dorman, et al. (1988) examined one-year self-reported crash rates of 127 IDDM individuals and their non-diabetic siblings. Although the overall crash risks of the two groups (7.1/100 drivers versus 14.2/100 drivers) were not significantly different, female diabetic drivers showed a marked increase for crashes. Compared to the female controls, female diabetic drivers had a five-fold increase in reported crashes. Overall, when adjusted for mileage driven, the number of crashes was higher in the IDDM population compared to the non-diabetic controls (10.4 versus 3.9 crashes per 100 drivers per 1 million miles), but this difference was not significant.

The following year, based on the results of a questionnaire mailed to individuals with Type 1 (IDDM) diabetes, Eadington and Frier (1989) compared the crash rates of 166 respondents to that of drivers in the general population. Mileage-adjusted crash rates for males were 4.9 crashes per million miles driven and 6.3 crashes per million miles for females, with an overall rate of 5.4 crashes per million miles driven. The authors concluded that the diabetics' crash rate was comparable to that of the overall population crash rate of 10 crashes per million miles driven (based on 1986 DOT statistics).

Stevens, Roberts, McKane, et al. (1989) compared the self-reported crash rates of 354 diabetic drivers treated with insulin to 302 non-diabetic outpatient drivers. The two groups showed similar characteristics in terms of annual distance driven and usual driving area. Crash rates were similar between the two groups with 23 percent of the diabetic drivers and 25 percent of the non-diabetic drivers reporting crashes in the previous five years. Importantly, as noted by the authors, the diabetic drivers included in this study, and those of Eadington and Frier (1989), were a select group of diabetic drivers. That is, individuals who had diabetic complications or difficulties with diabetes had often stopped driving, a consideration that may have contributed to the favourable crash record of the diabetic group. In Germany, a recent survey among 257 diabetic individuals treated with insulin revealed that study individuals reported a total of 27 severe car crashes during the previous two years (Chantelau, 1991). This translates into a rate of 0.06 severe crashes per driver per year, compared with approximately 0.07 such crashes per driver per year in the average population. Again, the major limitation is that risk exposure (e.g., annual mileage driven) was not taken into consideration in this investigation.

Finally, in a population-based, retrospective cohort study of 30,420 individuals (16 to 90 years of age), with and without diabetes mellitus or epilepsy, Hansotia and Broste (1991) compared the standardized rates of moving violations and crashes during a four-year period in affected and unaffected cohorts. The size of the final cohort of persons with diabetes included for study was 484. Standardized mishap ratios for individuals with diabetes for moving violations were non-significant, but the standardized crash ratio was 1.32 versus 1.00. Of interest, 90 percent of the diabetic individuals were considered to have Type II diabetes (NIDDM), although more than one third of the individuals of the NIDDM cohort took insulin. Again, the study is limited by the lack of information on the number of miles driven annually by study participants.

A summary of the available literature reveals a lack of consensus on the risk of crashes for diabetic drivers. Generally, the results from the earlier studies are based on crash and violation data obtained from state records while those from later years are, for the most part, based on data obtained from self-reports. Despite the limitations of each of the methodologies, there are, however, no clear-cut patterns either for time period or methodology. That is, two of the earlier studies (Crancer and McMurray, 1969; Waller, 1965) reported an increase in crashes among diabetic drivers, while Ysander (1966) and Davis et al. (1973) found no significant differences. Among the later studies, Hansotia and Broste (1991) found significant increases for diabetic drivers and Songer et al. (1988) found increases for female diabetic drivers. Eadington and Frier (1989), on the other hand, reported decreases in crash frequencies, while Stevens et al. (1989) found no differences. Five of the investigations utilized state records for determining crash risk, and in three of those (Crancer and McMurray, 1968; Hansotia and Broste, 1991; Waller, 1965), significant increases for diabetic drivers were found. However, a decrease was noted in Ysander's (1966) investigation, and no differences were reported in Davis et al.'s (1973) investigation.

Studies relying on self-report are equally disparate. One investigation found a decrease in frequency of crashes for diabetic drivers (Eadington and Frier, 1989), results from another indicated no difference (Stevens et al.1989), while the third (Songer et al., 1988) reported an increase but for female diabetic drivers only. The pattern of results is equivocal even when distinguishing between drivers with IDDM versus those in which type of diabetes is unspecified. Similarly, no clear-cut pattern emerges when amount of driving exposure is taken into consideration. However, when one considers country of origin, a pattern emerges. That is, studies done in the United States generally show either an increased crash risk or a trend toward increased crashes for individuals with diabetes mellitus (Crancer and McMurray, 1968; Hansotia and Broste, 1991; Koepsell, Wolf, McCloskey, et al., 1994; Songer et al., 1988; Waller, 1965). The European studies, on the other hand, fail to show a significant difference in crash rates. Thus, differences in licensing requirements may account for the differences in crash rates for the different countries. It may be that the European countries have had more restrictive licensing requirements and/or there may be more awareness in European countries of the associated risk than in America. Therefore, diabetic drivers in European countries may be compensating for their illness by driving less.