MORTALITY STATISTICS

A blogsite has reported that a river resident has lived on the Tittabawassee River for 45 years and has raised six healthy daughters.  She believes that this is ample confirmation that dioxins are not a risk to human health.   More than one Midland resident believes that dioxins are harmless for basically the same reason–no cancer in my family.

Last year, I had a conversation with a prior resident of Midland about a "cancer cluster" in his neighborhood.  In addition to his wife that died from colon cancer, he named at least ten neighbors that had been either diagnosed with cancer or had died from cancer.  Perhaps, his opinion about the health risk from dioxin exposure might be different than that of a certain river resident.

However, despite the "cancer cluster", there is no evidence that all dioxin exposure will result in cancer.  However, there is ample evidence that dioxin exposure is harmful to human health -- not to all those exposed, but many.

Cancer is a disease that has affected so many that a statistical analysis can be carried out with some degree of certainty.  While statistics can not predict who will contract cancer, there is ample data to determine if the incidence of cancer in larger populations is abnormally high or low.

Opponents that oppose controlling dioxin exposure to very low levels may be basing their opinion on the assumption that, unless there is a cancer "epidemic", dioxins are harmless.

In any cancer mortality study, there is a very fine line between a "statistically significant" linkage and a "no causal relationship".  It is amazing as to how few deaths are needed to make the difference between the two findings.

In 1991, the National Institute for Occupational Health and Safety (NIOSH) published a study that tracked the mortality of 5172 chemical workers that had been exposed to TCDD.  Dow's Midland plant dioxin exposed employees were included in the study and totaled approximately 2200 workers.  The study, Cancer Mortality in Workers Exposed to 2,3,7,8-Tetrachlorodibenzo-dioxin, Fingerhut, et al, 1991, examined the mortality experience of the 5172 male workers from 1942 to 1987.

All of the cancer specific information from the study can be found on a table that compares Standardized Mortality Ratio's (SMR) from the NIOSH study to six Dow mortality studies.  The link to the table is Link To NIOSH Study

Two findings from the NIOSH study are particularly interesting.

Cancer Mortality (a)

                                           Obs.    Exp.     SMR     95% CI

          All Cancers                265     229.9     115     102 - 130  (p<0.05)
          Cancers - Rare(b)         24       14.8     162     104 - 241  (p<0.05)

                  (a) Fingerhut, et al, 1991
                  (b) Cancers of Other and Unspecified Sites - no ICD category

Both findings are statistically significant as indicated by the lower 95% Confidence Interval (95% CI) value being greater than 100.

Please be aware that a very small number of additional deaths–35.1 (Obs. 265 versus Exp. 229.9)–were all that were necessary to make the All Cancers mortality ratio statistically significant. 

Now a cynical person may claim that an additional 35.1 deaths spread over 45 years over 5172 employees does seem like much–and it's not.  However, these additional deaths are very real to the families of the deceased and they are significant enough to confirm that dioxins are harmful to human health.

The study concluded, "An unexpected finding was the small but significant increase in mortality from all cancers combined.  The observed increase is consistent with a carcinogenic effect of TCDD.... Moreover, a significantly increased SMR for all cancers combined is unusual in occupational studies of chemical workers."

The NIOSH study also noted an statistically significant elevation of Cancers of Other Unspecified Sites.  The cancers that are included in this category are the cancers that are so rare and occur in such low numbers that they are not assigned a specific code number in the International Classification of Diseases.

For these rare cancers, the difference between 'statistically significant" and "no causal relationship" is even less than in All Cancers.  The NIOSH study reported the following findings for these rare cancers: Obs. 24, Exp. 14.8, SMR 162, 95% CI 104-241.  Only 9.2 additional deaths were found over 5172 workers over 45 years–enough to make the elevated number of deaths statistically significant.

The elevation in mortality from these rare cancers was probably not surprising to Dow since the company has observed similar results at the Midland plant site.

Dow Chemical Studies
Cancer Mortality - Other and Unspecified Sites

                                             Obs.        Exp.        SMR       95% CI

          Bond, 1987 (a)               169       125.3         135       115-157 (Total cohort)
          Bond, 1987 (a)                126        89.5         140        117-168 (Hourly)
          Ott, 1987                          12          4.6         261       135-456
          Cook, 1986                        9          3.7         243       111-462
          Cook, 1987                       12         4.6         261       135-456

                  (a) Study is not dioxin-specific–general chemica exposure only

Ott, 1987 had these comments on the findings, "Among other cancer sites, there was a significant increase in deaths due to other and unspecified malignant neoplasms (12 observed v 4.6 expected).  Review of the death certificates... did not provide additional insights as to the origin of the tumors. Nor did the internal analyses identify any significant work area or dixoin-related associations. Aside from noting that mortality for the category tends to run somewhat higher at this plant location [Midland]... we can offer no explanation for the finding."

Perhaps there is an explanation.  Since the NIOSH found a similar elevation in the 5172 workers from 12 different companies that had been exposed to dioxins, this suggests that the increased mortality is associated with dioxin exposure.

The Bond, 1987 study is especially significant since the study evaluated the mortality of approximately 37,682 male employees that worked at Dow's Midland and Bay City sites from 1940 to 1982.  Although the study did not provide specific mortality data for the Midland plant, it can be assumed that the 29,420 (78%) Midland plant employees endured the greatest amount of chemical exposure and probably contributed the greatest number of deaths.

This study also reported a statistically significant elevation in mortality from these rare cancers in both the total cohort, the hourly workforce and the salaried non-exempt workforce (Obs. 14 Exp. 6.6 SMR 213 (116-356).

It is well recognized that the Dow's Midland site is heavily contaminated with dioxins that were emitted from Dow's hazardous waste incinerators.  It is very probable that dioxin exposure is the reason for the statistically elevated mortality from these rare cancers.

In the Bond, 1987 study, the additional number of deaths from these rare cancers was 43.7 deaths spread over almost 38,000 male employee spread over 42 years.

I really do not know how many long term residents live on property adjacent to the Tittabawassee River–let's assume 2000 and let's also assume that the dioxin exposure is similar to that experienced by the workers in the NIOSH study.

In the 45 years that a certain resident lived on the river and raised six daughters, a total of 88.9 cancers would be expected in the 20 miles of river from Dow's plant to the Saginaw River.  According to the NIOSH study results, an excessive number of cancer deaths would be 102.4 deaths.

It is very difficult to imagine that this resident would notice an additional 13.5 deaths over roughly 45 years on 20 miles of river.

If there are 5000 long term residents, the additional number of cancer related deaths would be 33.5 additional deaths over 45 years on the 20 miles of river.

Increased mortality from dioxin exposure does not reveal itself as a cancer epidemic... dioxin takes its victims one by one over many years... sometimes so slowly that only an epidemiologist would notice. 

Only a few others really notice... the persons that died prematurely and their families.