PRELIMINARY HEALTH REVIEWS IN RANCHO CORDOVA, SACRAMENTO COUNTY, CALIFORNIA
RANCHO CORDOVA, SACRAMENTO COUNTY, CALIFORNIA
Responses to Comment
The initial draft of this health consultation was released to the site team after apresentation of the data at the site team meeting on December 11, 1997. The period ofwritten comment extended from that time until January 15, 1998. Only one formalcomment was received from the site team participants and was from Aerojet. That letterhas been reproduced and included below.
We have the following comment on the "Actions Planned" section which suggests the possibility of exposure modeling in regard to a possible study of neonatal thyroid levels. The CDHS draft demonstrates that neonatal hypothyroidism was not increased in incidence in the areas of interest. In the "unexposed" area, CDHS found only four cases of neonatal hypothyroidism reported in twelve years, so there is little likelihood of doing any meaningful epidemiologic study of hypothyroidism, given so few cases available. Undertaking such a study to further evaluate neonatal thyroid hormone levels with mothers divided into groups according to maternal perchlorate intake as modeled byATSDR raises many methodological concerns, including:
1. There is uncertainty as to the accuracy and validity of the exposure modeling as a means to distinguish "exposed" from "unexposed" groups, and no way to measure the validity. We believe that an exposure assessment would be very speculative. We refer you generally to Aerojet's comments in our letter of October 13, relative to exposure assumptions. Your October 16 draft consultation also notes difficulties with attempting such an evaluation. Even if one could accurately model past concentrations in particular wells, there must be adequate information on consumption and a host of other factorsto make the evaluation potentially useful.
2. It is difficult to interpret differences in mean thyroid hormone levels inpopulations if almost all of the values are within normal limits and there is no excessnumber of persons with clinically significant abnormal values.
3. The consultation document points out the importance of other factors, such asdeficiencies of Thyrotropin Releasing Factor and Thyroid Stimulating Hormone, aplasia orhypoplasia of the thyroid gland, and iodine deficiency. The CDHS has not indicated howany potential study will deal with these other factors.
4. Finally, the utility and interpretation that may be given to population differencesin thyroid hormone levels when such levels are still within normal bounds is questionable.
On page one, paragraph two, Aerojet comments, "The October 16, 1997 draft states a belief that perchlorate first contaminated wells in the Cordova system "as early as 1987"which is an assumption which we think should not be stated as such.".
To the best of our understanding, perchlorate may have first contaminated drinking water wells in this area sometime during the late 1980s. The objectionable statement referring to 1987 was cited from a prior health consultation performed by ATSDR in 1993. We have not been given information to contradict this statement, but have elected to change the statement to reflect the possible uncertainty in the date of first contamination. This statement will now read "It is believed that perchlorate first contaminated wells in the Arden Cordovasystem sometime in the late 1980s."
In the following line also on page one, Aerojet comments, "We are also concerned about the statement at page 3 that water "may have posed a health hazard" and the absence of any language that notes that CDHS believes that health impact was unlikely. (See Aerojet's October 13 letter at pages8 and 9.)".
We see no reason to amend our assertion that water containing perchlorate may haveposed a health hazard. Water free of this substance would most certainly not pose ahealth hazard (from perchlorate). However, levels of perchlorate found in several wellsin the Arden Cordova water district were substantially above the USEPA's acceptableprovisional reference dose range (4-18 ppb). On the basis of this data we believe that thestatement is accurate and that the water indeed may have posed a health hazard. We didnot feel that any inclusion of language implying that the health department believed thathealth impact was unlikely was warranted because not all aspects of perchlorate-relatedhealth outcomes could be measured by our preliminary health outcome studies and thosethat were studied were extensively commented upon in the appropriate discussion section.
Based on the studies' preliminary results wherein no significant health effects werefound in the potentially affected population, and the uncertainties of how other factorswould also impact the observable health impacts, it appears reasonable to conclude thatmodeling of potential perchlorate uptake is not warranted. However, if CDHS intends toundertake the exposure evaluation described, it is our understanding Aerojet's input will beincluded in the evaluation of methods to be used and developing the assumptions andapproach to be taken.
We were asked by parties during the presentation of this data at an Aerojet SiteAssessment Team meeting in Rancho Cordova on December 11, 1997 to rerun our dataon neonatal hypothyroid cases and restrict the suspected exposure period from 1990through 1995 or 1996 and the affected zip code to 95670 only. This seemed reasonablesince there was some consensus in this meeting that the estimated time of exposure wastoo broad and the area of exposure was more limited than we previously thought. I haveincluded the results of the original analyses and the additional analyses below forcomparison:
|Exposure Status/Location||Hypothyroid Cases |
|Rancho Cordova (exposed)*||4 (3.76)||11,814|
|Rancho Cordova (unexposed)**||6 (6.41)||20,135|
|Rest of State (unexposed)||2,068 (-)||6,494,193|
|Exposure Status/Location||Hypothyroid Cases |
|95670 (exposed)||4 (1.84)||5,217|
|Rest of State (unexposed)||1391 (-)||3,954,281|
|Odds Ratio = 2.18 |
95% Confidence Interval ( 0.85 - 5.6)
Fisher's Exact p-value = 0.115
As can be seen from the above tables, the results are different depending on when it is estimated that individuals consumed potentially perchlorate-contaminated water. If we consider the data from table 1a, then in fact, individuals were shown to be over two times more likely to have given birth to a child with confirmed clinical hypothyroidism with maternal residence in zip code 95670 during 1990-1996 as compared to births throughout the state. However, these results were not statistically significant and may in fact be unrelated to perchlorate exposure.This information does make it more important to follow-through with an exposure assessment for perchlorate contamination of the area's drinking water supply, which will in turn allow for improved analyses of the relationship of drinking water and health parameters.
There are four sections on page two which deal with methodological difficulties in designing a studywhich would measure the effects of maternal perchlorate ingestion (via drinking water) and neonatalthyroid hormone status.
We agree with the comments provided by Aerojet regarding the potential difficulties and limitations ofexposure modeling, confounding factors and interpretation of results. We do not feel that thesedifficulties should prevent us from attempting to perform these studies however, but they must and willbe considered in the designing of the study and the interpretation of any results which are ultimatelyobtained. We look forward to further comments and suggestions from Aerojet regarding futureevaluations.
|Exposure Status||Number |
|Rest of State||6,479,190||14.77||14.4||4.38|
|Exposure Status||Number |
|Rest of State||2,851,284||15.03||14.7||4.56|
Figure 2. Rancho Cordova Zipcode 95670 and its Component Census Tracts
1 The OSHPD maintains a computerized data base of all discharges to California hospitals for each year. Pertinent patient information found in this data base includes: primary diagnosis with listings of additional diagnoses, demographics and treatments performed while hospitalized.
2 The CCR is the statewide population-based cancer surveillance system which gathers information on all cancers diagnosed in California since 1988 (excluding basal and squamous cell carcinomas of the skin and carcinoma in situ of the cervix). Hospitals and physicians are required by law to refer information to the CCR's network of ten regional registries which perform analyses and conduct studies on this data.