ON THE RISK September 2024

Table 3. Results of Strategy 3.

Women

Men

Obs

Exp

MR(Obs/Exp)

Obs

Exp

MR(Obs/Exp)

Neither 0.0004 0.0015

0.24 0.24 0.49 0.44

0.0006 0.0022 0.0012 0.0051 0.0022 0.0047 0.0035 0.0104

0.26 0.24 0.48 0.33

Smoker 0.0010

0.0042

Diabetic 0.0017 0.0034

Both

0.0034 0.0079

RMR: Diabetes vs. No Diabetes Among Non-Smokers:

2.07 1.78

1.81 1.40

Among Smokers:

Note : Observed and expected mortality rates are 10-year geometric average annual mortality rates.

non-smokers (diabetic non-smokers vs. non-diabetic non-smokers) against its impact in smokers (diabetic smokers vs. non-diabetic smokers). In both men and women, we see that the impact of diabetes in smok- ers is proportionately less than it is in non-smokers. Discussion This study demonstrated something interesting– that the effect of diabetes on smokers is, when expressed as a ratio, less than that in non-smokers. This may be unexpected; however, it is also true that if the dif - ference is expressed as excess death rate instead of a mortality ratio, then the impact of diabetes is greater in smokers. For example, in women the excess death rate for diabetes among non-smokers is 1.3 deaths

per thousand per year, while it is 2.4 for diabetes among smokers. This is because the baseline mortal- ity risk in smokers is already quite high, and therefore diabetes has a hard time matching or exceeding that impact. The likely reason we are seeing this evidence that combo debits are not warranted is that smok- ing is treated in a unique way in life insurance. It is treated as a mortality multiple that varies with age, sex and policy duration. These smoker adjustments are fairly large, so when a rating is applied to a smoker because they have diabetes, this table rating is multiplying an already elevated mortality. Figure 1 shows that, for the peak insurance-buying ages of 30-60, the average mortality multiple for smoking is between 2.5 and 3.0 until the very late durations.

Figure 1. 2015 VBT Smoker Hazard Ratios. Average of Ages 30-60, by Duration.

ON THE RISK vol.40 n.3 (2024)

72

Powered by