IH5426_Gutschmidt_J_EPI Paper Final Project

1

Jeffrey Gutschmidt, December 5, 2016 – IH5426: Advanced Epidemiology – Critical

Review of: Occurrence of mental illness following prenatal and early childhood

exposure to tetrachloroethylene (PCE)-contaminated drinking water: a retrospective

cohort

Critical Review of:

Occurrence of mental illness following prenatal and early childhood exposure to

tetrachloroethylene (PCE)-contaminated drinking water: a retrospective cohort

by: Jeffrey A Gutschmidt

For: Advanced Epidemiology, Fall 2016

At: Montana Tech of the University of Montana

Presented to: Professor Theresa Stack

Author Note

11418 NE 128th

St #58, Kirkland, WA 98034: (206) 325-3446

2


Review

Table of Contents

Background .......................................................................................... 5

Historic Usage of Tetrachloroethylene.................................. 5

Suspected or Known Human Disease Outcomes ................... 6

Study Description ................................................................................. 8

Study Rationale ................................................................... 8

Study Subjects. .................................................................... 9

Study Interviews/Questionnaire. ........................................ 10

Study Analysis. .................................................................. 12

Study Results. ................................................................... 13

Study Limitations. ............................................................. 13

Discussion .......................................................................................... 14

Study Strengths. ................................................................ 14

Study Weaknesses. ............................................................ 14

Study Hypothesis............................................................... 14

Study Bias and Error. ......................................................... 15

Study Findings. ................................................................. 16

Study Limitations. ............................................................. 17

Study Comparisons. .......................................................... 17

3


Review

Conclusion .......................................................................................... 18

References .......................................................................................... 19

Tables ................................................................................................. 21

Table 1 ............................................................................. 21

Table 2 ............................................................................. 22

Table 3 ............................................................................. 25

Figures ............................................................................................... 26

Figure 1 Subjects by Mental Health Outcomes. ................... 26

Figure 2 Initial Selection. ................................................... 26

Figure 3 Final Selection ..................................................... 27

Figure 4 Subjects by Status and Birth Year ........................ 27

Figure 5 Subjects by Status, Gender, and Race ................... 28

Figure 6 Subjects Available for Analysis ............................. 28

Figure 7 Subjects by Education Status. ............................... 29

Figure 8 Subjects by Employment Status ............................ 29

Figure 9 Subjects by Marital Status .................................... 30

Figure 10 Subjects by History of Major Illness. ................... 30

Figure 11 Subjects by Solvent Exposure in Employment ...... 31

Figure 12 Subjects by Hobbies with Solvent Exposure. ........ 31

4


Review

Figure 13 Age of Parents at Subjects Birth ......................... 32

Figure 14 Mother’s Education Level at Subjects Birth .......... 32

Figure 15 Father’s Occupation at Subjects Birth ................. 33

Figure 16 Mother’s Prenatal Care at Subjects Birth ............ 33

Figure 17 Mother’s Smoking Status at Subjects Birth .......... 34

Figure 18 Mother’s Alcohol Consumption at Subjects Birth . 34

Figure 19 Mother’s Marijuana Usage at Subjects Birth ....... 35

Figure 20 Mother’s Medical Complications ......................... 35

Figure 21 Mother’s Occupational Solvent Exposure ............ 36

Figure 22 Subject’s Family History of Schizophrenia .......... 36

Figure 23 Subject’s Family History of PTSD. ....................... 37

Figure 24 Subject’s Family History of Bipolar Disorder ....... 37

Figure 25 Subject’s Family History of Depression ............... 38

End Notes ........................................................................................... 39

5


Review

Background

Tetrachloroethylene (PCE) is a solvent that is manufactured and used in dry-

cleaning, degreasing, and in the manufacture of other chemicals (New York

State, 2016). It is found in consumer products such as paint, spot removers,

water repellents, brake and wood cleaners, glues, and suede protectors (New

York State, 2016). Tetrachloroethylene is also known by the following

synonyms: PERC, perchloroethylene, and PCE (New York State, 2016).

Historic Usage of Tetrachloroethylene

PCE was first used in the dry-cleaning industry in 1964 (State Coalition for

Remediation of Drycleaning, 2007). The dry cleaning industry is responsible for

90% of PCE usage in the U.S. (State Coalition for Remediation of Drycleaning,

2007). Prior to PCE, Stoddard Solvents had been in use since 1925 (Whittaker,

2013). However, the flash point of Stoddard Solvents being 104 degrees

Fahrenheit was problematic for the industry due to the risk of fires. Non-

flammable PCE became the preferred solvent due to safety concerns with

Stoddard Solvent and its increased effectiveness as a dry-cleaning solvent for

clothing (Whittaker, 2013). Production of PCE in the U.S. is declining. As of

2000 (most recent data available), only three companies in the US produced

PCE: Dow Chemical, PPG Industries, and Vulcan Chemicals (Newcombe, 2000).

Aside from dry cleaning, the next major industrial user of PCE is the degreasing

industry, which uses 80% of the non-dry cleaning PCE produced (Newcombe,

2000).

6


Review

Historic poor waste disposal practices and releases at dry cleaning

facilities nationwide have led to legacy pollution at an estimated 70% of the

more than 35,000 sites nationally (Keys, 1998). “PCE is heavier than water.

When spilled, it swiftly sinks, penetrating concrete, soil and practically anything

else in its way. When PCE reaches groundwater, it does not dissolve. Instead,

PCE can collect against bedrock, forming pools that remain volatile for decades

if the chemical is not removed or neutralized with other chemicals. Once in soil

and groundwater, PCE can resurface as toxic vapors seeping into buildings”

(Finley, 2016, p. 1). Accordingly, PCE is persistenti

when released in the

environment, is a major source of groundwater contamination in the U.S., and is

difficult to clean up. It may come from illicit disposal or leaking sanitary sewer

lines, in which municipal sewage authorities at one time permitted dry cleaning

wastes to be disposed (Keys, 1998). Today, most cleanup sites are identified

when property changes ownership, and banks perform Environmental Site

Assessmentsii

prior to loan approval (Finley, 2016).

Suspected or Known Human Disease Outcomes

According to Newcombe (2000) “In humans, PCE is known to cause toxic

effects in the liver, kidney, and central nervous system. It is suspected to cause

harm to reproductive and developmental health, and is a possible carcinogen.

The major human exposure route is inhalation of contaminated air” (p. 1) but

can also include ingestion of contaminated drinking water mainly form private

and municipal wells due to lack of testing and treatment (Finley, 2016). The U.S.

7


Review

Environmental Protection Agency (EPA) sets the maximum contaminant level

(MCL) for PCE in drinking water at 5.00 µg/L based on what EPA calls observed

acuteiii

health effects from PCE above the MCL, which include detrimental effects

to the liver, kidney, and central nervous system (United States Environmental

Protection Agency, 1995).

8


Review

Study Description

This retrospective cohort study looks at the relationship between prenatal

and early childhood exposure to tetrachloroethylene (PCE) in drinking water and

an increased risk of depression, bipolar disorder, post-traumatic stress

disorder, and schizophrenia among adults from the Cape Cod region of

Massachusetts (Aschengrau, et al., 2012).

Study Rationale

From the 1960s through the early 1980s public water companies in

Massachusetts installed vinyl lined asbestos cement (VL/AC) water pipes to

correct issues with high pH drinking water. These pipes were sprayed with PCE.

Tests in 1980 showed that large quantities of PCE were leaching from these

pipes into the drinking water of affected residents (Ashengrau, et al., 2012).

This study attempts to look at one set of possible human health effects from

these known exposures.

Drinking water levels in Cape Cod, one of the affected communities, were

measured at a range of 1.5 to 7,500 µg/L. Previously published research (see

Study Comparison) has linked PCE exposure in children of dry cleaners to a 3.4-

fold increase in the relative risk of developing schizophrenia (Aschengrau, et

al., 2012). Following the 1980 tests, pipes were flushed with the goal of

reducing PCE levels in drinking water to below 40 µg/L, the suggested action

level at the time, whereas the modern MCL, per U.S. EPA guidelines for PCE, is 5

9


Review

µg/L (United States Environmental Protection Agency, 1995)/ (Aschengrau, et

al., 2012).

Study Subjects.

Eligible subjects were selected from persons born to married women who

lived in the identified Cape Cod towns with affected water supplies and who

were born from 1969 through 1983 (Aschengrau, et al., 2012). Maternal

addresses were cross-matched with utility records that showed the location and

installation year of the VL/AC pipes (Aschengrau, et al., 2012). Based on this

information, two index groups were initially selected based on exposure status

(N=1,910) (Aschengrau, et al., 2012), (Figure 2). Additionally, another 1,202

subjects were identified as being older siblings of the index subjects who were

also born between 1969 and 1983 and they were initially considered to be

unexposed because they were born prior to the family moving to the affected

address. Birth records for each study subject were examined including the

name of the subject, name of the parents, birth date (Figure 4), birth weight,

gestational duration, and the parents’ age (Figure 13) and education level

(Figure 14) at the time of birth (Aschengrau, et al., 2012).

Next, study subjects were traced to obtain addresses and phone

numbers. Letters were sent to successfully traced subjects, and a questionnaire

was included in the letter with a request that it be completed and returned

(Aschengrau, et al., 2012), (Figure 3).

10


Review

The final group of subjects consisted of 1,512 (Table 1) persons, of

whom 585 were exposed index subjects, 562 were unexposed index subjects

and 365 were unexposed older siblings of index subjects (Aschengrau, et al.,

2012). However, during the detailed exposure assessment that followed, 414

subjects (27.4%) changed status from exposed to non-exposed (Aschengrau, et

al., 2012), (Figure 6). The demographic make-up of study subjects was

predominantly college educated women in their late twenties who were

employed and either married or cohabitating with a partner when they

completed the questionnaire (Aschengrau, et al., 2012), (Figure 5).

Study Interviews/Questionnaire.

According to Aschengrau (2012):

A self-administered questionnaire was sent to all successfully

traced subjects to gather information on each subject’s demographic

characteristics including race, ethnicity, current marital status [(Figure 9)],

educational level [( Figure 7 )], and occupation [(Figure 8)]; personal

history of chronic illnesses [(Figure 10)] and obstetrical and gynecological

problems [(Figure 20)]; family history of mental disorders [ (Figures 22-

25)] occupational and non-occupational sources of solvent exposure

[(Figures 11-12)] and residences from birth through 1990, including the

exact street address and calendar years of residence for all Cape Cod

addresses. Information was also collected on the occurrence of four

mental illnesses: depression, schizophrenia, bipolar disorder or manic-

11


Review

depressive disorder, and post-traumatic stress disorder (PTSD) [( Figure

1)]. In particular, subjects were asked if a doctor or health care provider

ever said that they had the illness, and if the subjects responded

affirmatively, they were asked for the year of diagnosis. Lastly, the survey

gathered information on the subject’s knowledge of the PCE

contamination and self-assessments of their PCE exposure [(Table 2)].

Additional data were also available for 81% of subjects whose

mothers participated in our [Ann Aschengrau et al.] prior study on the

impact of PCE exposure on reproduction and development. These data

included maternal characteristics, such as changes in marital status;

cigarette smoking [(Figure 17)], alcoholic beverage consumption [(Figure

18)], marijuana use [(Figure 19)], medical and obstetrical complications,

and prenatal care during the subject’s gestation [(Figure 16)]; breast

feeding practices; and exposure to solvents[( Figure 21 )]. We attempted

to collect information on the mother’s water consumption and bathing

habits during the subject’s gestation and the subject’s consumption and

bathing habits during childhood, but this information could not be

recalled by a sizable portion of women and the prevalence of key

variables was relatively low. Thus, these data were not incorporated into

our [Ann Aschengrau et al.] present analyses. (p. 9)

12


Review

Study Analysis.

Relative risk ratios (RR) were used to estimate the strength of association

between exposure and disease. Crude analysis was performed and then a

statistical model called GEE (generalized estimating equationiv

) was applied to

account for non-independent outcomes from several children in the same

family. Of the subjects of interest 39% were siblings (Aschengrau, et al., 2012).

A confidence interval of 95% was established for the evaluation of relative risk

ratio data (Aschengrau, et al., 2012).

Confounding was addressed with adjusted GEE analysis. This analysis

considered the following criteria “gender, race, age, birth weight, gestational

duration, history of chronic illnesses and obstetric/gynecologic problems

predating the occurrence of any mental illness, and history of solvent-related

jobs and hobbies; family history of mental illness (including depression,

schizophrenia, bipolar disease and post-traumatic stress disorder among first

degree relatives); the mother’s age and educational level when the subject was

born; paternal age, educational level and occupation when the subject was

born; the mother’s prenatal care, multivitamin use, alcoholic beverage

consumption, cigarette smoking, marijuana use, medical conditions, and

obstetrical complications when she was pregnant with the subject;

breastfeeding history; number of live-born siblings and sibling deaths; and

maternal history of solvent exposure” (Aschengrau, 2012, p. 12).

13


Review

Study Results.

Twenty-two percent of the study subjects reported having at least one

mental illness. The breakdown of these illnesses was 17.5% depression, 3.7%

bipolar disorder, 4.8% post-traumatic stress disorder, and 0.03% schizophrenia.

Four percent of the subjects reported having more than one mental illness.

Increase in risk of depression was not reported. However, increased relative risk

scores were observed in bipolar disorder at a 1.8-fold increase, post-traumatic

stress disorder a 1.5-fold increase, schizophrenia a 2.1-fold increase (Table 3)

(Aschengrau, et al., 2012).

Study Limitations.

Exposure misclassification is likely an issue with this study. Reporting

bias is also suspect (see Study Bias and Error). The prevalence of illness in the

study group was lower than that observed in the general population surveys of

the United States: 2001-2003, lifetime prevalence of depression 22.7%, dipolar

disorder 4.5%, and PTSD 8.2%. based on a nationally representative sample of

30-44-year-old adults (Aschengrau, et al., 2012).

14


Review

Discussion

The following sections comprise a critical review of this study.

Study Strengths.

This study clearly demonstrates an association between exposure to PCE

and some increased risk of negative mental health outcomes, especially bipolar

disorder and PTSD.

Study Weaknesses.

The demonstrated associations are not as strong as they might be due to

likely errors in classification of the study participants (exposed versus non-

exposed). For this reason, “the relative risk or odds ratio tends to be diluted”

(Gordis, 2014, p. 7598).

Study Hypothesis.

Specifically, no formal statement of a hypothesis was made, as such. Nor

was a stated null-hypothesis tested. However, in the opening lines of the study

paper the following statement is made, which could be taken as the study

hypothesis: “This retrospective cohort study examined whether early life

exposure to PCE contaminated drinking water influenced the occurrence of

depression, bipolar disorder, post-traumatic stress disorder, and schizophrenia

among adults from Cape Cod, Massachusetts" (Aschengrau, et al., 2012, p. 4).

15


Review

Study Bias and Error

Exposure misclassification, a form of nondifferential misclassification which

tends to dilute the observed relative risk to a score closer to parity with one

(Gordis, 2014), may be an issue with this study. This is suspected because,

for instance, the association (relative risk) of exposure to PCE and

schizophrenia was much lower than the rate reported in the Perrin study (2.1

vs. 3.4) (see Study Comparisons). Historical exposure measurements were not

available and estimates were used in place of real data based on EPANET water

distribution software (Aschengrau, et al., 2012). Additionally, the study

participants were highly weighted toward female over male (60.7% vs. 43.5%),

more likely to have a more highly educated mother (61.0% vs. 49.3%),

predominantly white, over 98%, and more likely to have a mother who

participated in a previous study (81%) conducted by the same authors of this

study (Aschengrau, et al., 2012). Taken together, this indicates a possibility for

selection bias, though the author points out that an equal proportion of

exposed and non-exposed subjects had similar characteristics (Aschengrau, et

al., 2012). The participant pool disproportionately came from a more highly

educated portion of the population. Gender and racial bias are possible because

the study is highly skewed toward white female respondents. Recall and/or

reporting bias are possible (Gordis, 2014) because the data on mental health

outcomes was based on self-reports (Aschengrau, et al., 2012). The accuracy of

self-reports may be affected by lack of knowledge (recall bias) of actual health

16


Review

outcomes or an underreporting due to the social stigmasv

associated with

mental illness (reporting bias). Note that the lifetime prevalence rates of mental

illness reported in this study were lower than rates reported nationally for the

age group of 30-44 year-old adults from 2001-2003 (Aschengrau, et al., 2012).

Additionally, the study authors reported that confounding errors may be

present in the study design due to “missing data on several risk factors for

mental illness” (Aschengrau, et al., 2012, p. 16). Finally, a low response rate of

eligible participants reduced the overall statistical power (1-β) of the study - a

type II error which can cause a false conclusion that treatments (exposed vs.

unexposed) do not differ, when in fact they do (Gordis, 2014).

Study Findings.

The study concluded that subjects with prenatal and early childhood

exposure to PCE had a 1.8-fold increased relative risk or bipolar disorder and a

1.5-fold increased relative risk of post-traumatic stress disorder (Aschengrau, et

al., 2012). The study showed an increased relative risk for schizophrenia of 2.1

(Figure 1) but the sample size of persons reporting this disorder was too small

to make meaningful inferences. Moreover, a greater association of risk of

bipolar disorder and post-traumatic stress disorder was demonstrated amongst

participants who were shown to be more highly exposed to PCE with relative

risk scores of 2.7 and 1.7 respectfully (Aschengrau, et al., 2012).

17


Review

Study Limitations.

This study was limited by low response rates (Aschengrau, et al., 2012),

misclassification bias, gender and race selection bias, and recall and reporting

bias from self-reporting (Aschengrau, et al., 2012). Together, these factors led

to reduced statistical power (Gordis, 2014) of the study and likely differential

misclassification (Gordis, 2014) which probably reduced the observed effects of

prenatal and early childhood exposure to PCE.

Study Comparisons.

A similar cohort study by Perrin et al in 2007 showed a clear connection

between exposure to PCE in dry cleaners and risk of schizophrenia in their

offspring (Perrin, et al., 2007). In Perrin et al., 88,829 live offspring were

followed from 1998 onward. Of these persons, 637 were admitted to the

hospital for schizophrenia (Perrin, et al., 2007). A relative risk ratio (Gordis,

2014) was calculated that showed a 3.4 times increased risk of schizophrenia

among children born to parents who were dry cleaners versus children whose

parents had some other occupation. The confidence interval of this study was

95% (1.3-9.2 range, p=0.01) and the results were controlled for by parental age,

a recognized risk factor for schizophrenia (Perrin, et al., 2007).

18


Review

Conclusion

In conclusion, more research is needed to draw a clearer connection

between prenatal and early childhood exposure to PCE and bipolar disorder,

post-traumatic stress disorder, and schizophrenia. However, the population of

exposed persons from the Cape Cod area, is an excellent population source for

this type of research because of its size and the length and magnitude of its

PCE exposure. Further research in this area could lead to more sensible policies

for PCE use in the workplace and other settings. Human exposure to PCE can

come from exposure to vapors, physical contact, or through exposure to

contaminated drinking water supplies. Additional research has the potential to

uncover and possibly mitigate negative health outcomes in exposed

populations. Lessons learned from this study, that would benefit future efforts

to study the effects of PCE exposure, include, the importance of trying to

achieve higher response rates—perhaps through door-to-door survey gathering

techniques—and controlling for reporting or recall bias through, mechanisms

such as, in-person interviews to more clearly establish mental health outcomes

among respondents. Unfortunately, both measures would likely lead to

significant increased cost.

19


Review

References

Aschengrau, A., Weinberg, J. M., Janulewicz, P. A., Romano, M. E., Gallagher, L.

G., Winter, M. R., . . . Ozonoff, D. M. (2012, January 20). Occurance of

mental illness following prenatal and early childhood exposure to

tetrachloroethylene (PCE)-contaminated drinking water: a retrospective

cohort study. Environmental Health, 32. Retrieved November 6, 2016,

from http://www.ehjournal.net/content/11/1/2

Finley, B. (2016, November 11). PCE Cleanup Sites. The Denver Post, p. 1.

Retrieved from http://extras.denverpost.com/pce/pcemobile.html

Gordis, L. (2014). Epidemiology [Kindle Windows 8 version]. Retrieved from

Amazon.com (Fifth ed.). Philidelphia, Pennsylvania: Elsevier Saunders.

Keys, G. (1998, September/October). CCIM Institute. Retrieved from Cleaning

Up After Dry Cleaners: http://www.ccim.com/cire-

magazine/articles/cleaning-after-dry-cleaners/?gmSsoPc=1

New York State. (2016, November 6). Department of Health. Retrieved from Fact

Sheet: Tetrachloroethene (PERC) in Indoor & Outdoor Air:

http://www.health.ny.gov/environmental/chemicals/tetrachloroethene/

Newcombe, R. (2000). Perchloroethylene (PCE) In Dry Cleaning Establishments.

Boise: University of Idaho. Retrieved November 6, 2016, from

http://www.webpages.uidaho.edu/etox/resources/case_studies/pce_dry.

pdf

http://www.ehjournal.net/content/11/1/2

http://extras.denverpost.com/pce/pcemobile.html

http://www.ccim.com/cire-magazine/articles/cleaning-after-dry-cleaners/?gmSsoPc=1

http://www.ccim.com/cire-magazine/articles/cleaning-after-dry-cleaners/?gmSsoPc=1

http://www.health.ny.gov/environmental/chemicals/tetrachloroethene/

http://www.webpages.uidaho.edu/etox/resources/case_studies/pce_dry.pdf

http://www.webpages.uidaho.edu/etox/resources/case_studies/pce_dry.pdf

20


Review

Perrin, M. C., Opler, M. G., Harlap, S., Harkavy-Friedman, J., Kleinhaus, K.,

Nahon, D., . . . Malaspina, D. (2007). Tetrachloroethylene exposure and

risk of schizophrenia: offspring of dry cleaners in a population birth

cohort, preliminary findings. National Institute of Health, 5.

State Coalition for Remediation of Drycleaning. (2007, November). Retrieved

from https://drycleancoalition.org/download/drycleaning-

historical_developments.pdf

United States Environmental Protection Agency. (1995, October). National

Primary Drinking Water Regulations: Tetrachloroethylene. Retrieved from

EPA: https://nepis.epa.gov/Exe/tiff2png.cgi/600012BY.PNG?-r+75+-

g+7+D%3A%5CZYFILES%5CINDEX%20DATA%5C95THRU99%5CTIFF%5C000

01939%5C600012BY.TIF

Whittaker, S. G. (2013). Evaluation of Solvon K4 in an Acute Fish Toxicity Test.

Seattle, WA: King County Local Hazardous Waste Management Program.

Retrieved November 6, 2016, from

http://www.hazwastehelp.org/publications/eDownload.aspx?DocID=FZt0

M%2fGXItE%3d

https://drycleancoalition.org/download/drycleaning-historical_developments.pdf

https://drycleancoalition.org/download/drycleaning-historical_developments.pdf

https://nepis.epa.gov/Exe/tiff2png.cgi/600012BY.PNG?-r+75+-g+7+D%3A%5CZYFILES%5CINDEX%20DATA%5C95THRU99%5CTIFF%5C00001939%5C600012BY.TIF



http://www.hazwastehelp.org/publications/eDownload.aspx?DocID=FZt0M%2fGXItE%3d

http://www.hazwastehelp.org/publications/eDownload.aspx?DocID=FZt0M%2fGXItE%3d

21


Review

Tables

Table 1

Table 1:© 2012 Aschengrau et al.; licensee BioMed Central Ltd. This is an open access

article distributed under the terms of the Creative Commons Attribution License

(http://creativecommons.org/licenses/by/2.0), which permits unrestricted use,

distribution, and reproduction in any medium, provided the original work is properly

cited.

Older Siblings

Exposed Unexposed Unexposed Total

Selected 1,910 1,928 1202 5,040

Excluded during

enrollment

Deceased 35 40 36 111

Parent refused

participation 148 80 427

Never located 113 149 70 332

No response 871 887 536 2,294

Refused 73 78 36 187

Returned

questionaire 619 626 444 1,689

Percent of selected 32.40% 32.5% 36.90% 33.5%

Percent Located 39.60% 39.3% 43.70% 40.5%

Excluded during

exposure

assessment

Inadequate

residential history 15 37 29 81

Off-Cape address in

town with VL/AC

pipes 19 27 50 96

Available for

analysis 585 562 365 1,512

Percent of selected 30.6% 29.1% 30.4% 30.0%

Final exposure

status

Both prenatal and

early childhood

exposure 561 160 110 831

Only early

childhood exposure 7 42 85 134

Unexposed 17 360 170 547

Total 585 562 365 1,512

Table 1

Index Subjects

Initial, Enrollment, and Initial and Final Exposure Status of Study Subjects

http://creativecommons.org/licenses/by/2.0

22


Review

Table 2





cited.

Characteristic

Year of birth n %/mean (sd) n %/mean (sd)

1969-1974 166 20.0% 131 23.9%

1975-1980 435 52.3% 288 52.7%

1981-1983 230 27.7% 128 23.4%

Current age (n,

mean, sd) 831 29.2 3.6 547 29.6 3.8

Gender

Male 331 39.8% 216 39.5%

Female 500 60.2% 331 60.5%

White race 818 98.4% 539 98.5%

Current educational

level

High School

graduate 128 15.4% 67 12.2%

Some College 192 23.1% 144 26.3%

Four-year college

grad or higher 510 61.4% 335 61.2%

Missing 1 0.1% 1 0.2%

Currently employed

Yes 719 86.5% 487 89.0%

No 92 11.1% 54 9.9%

MIssing 20 2.4% 6 1.1%

Current marital

status

Single 272 32.7% 157 28.7%

Married or

cohabitating 536 64.5% 371 67.8%

Other 19 2.3% 12 2.2%

Missing 4 0.5% 7 1.3%

Histroy of chronic

illness

Yes 111 13.4% 73 13.3%

No 644 77.5% 430 78.6%

Missing 76 9.1% 44 8.0%

Ever had job with

solvent exposure

Yes 123 14.8% 71 13.0%

No 687 82.7% 461 84.3%

Missing 21 2.5% 15 2.7%

Ever had hobby

with solvent

exposure

Yes 700 84.2% 462 84.5%

No 124 14.9% 79 14.4%

Missing 7 0.8% 6 1.1%

Mother's age at

subject's birth 831 27.2 4.7 547 27.5 4.4

Father's age at

subjects birth 831 29.8 5.7 547 29.8 5.3

Both Prenatal and Early Childhood Exposure Unexposed

Table 2


23


Review

Characteristic

Mother's Education

level at subjects

birth n %/mean (sd) n %/mean (sd)

High school

graduate or less 327 39.4% 178 32.5%

Some college 243 29.2% 188 34.4%

Four year college

grad or higher 260 31.3% 180 32.9%

Missing 1 0.1% 1 0.2%

Father's occupation

at subjects birth %/mean (sd) n %/mean

White collar 420 50.5% 257 47.0%

Blue collar 275 33.1% 170 31.1%

Other 126 15.2% 112 20.5%

Missing 10 1.2% 8 1.5%

Mother received

prenatal care during

subjects gestation

Yes 794 95.5% 520 95.1%

No 4 0.5% 0 0.0%

Missing 33 4.0% 27 4.9%

Mother smoked

cigarettes during

subject's gestation

11 + cigarettes a day 108 13.0% 59 10.8%

10 or fewer

cigarettes a day 74 8.9% 54 9.9%

None 483 58.1% 330 60.3%

Missing 166 20.0% 104 19.0%

Mother consumed

alcohol during

subjects gestation

1 + drinks a week 109 13.1% 76 13.9%

1-3 drinks a month 193 23.2% 125 22.9%

None 361 43.4% 242 44.2%

Missing 168 20.2% 104 19.0%

Table 2


24


Review

Characteristic

Mother smoked

marijuana during

subject's gestation n %/mean (sd) n %/mean (sd)

Yes 25 3.0% 18 3.3%

No 640 77.0% 420 76.8%

Missing 166 20.0% 109 19.9%

Mother had

medical and

obstetrical

complications

during subject's

gestation

Yes 122 14.7% 108 19.7%

No 536 64.5% 331 60.5%

Missing 173 20.8% 108 19.7%

Mother exposed to

solvents at work

Yes 76 9.1% 51 9.3%

No 573 69.0% 381 69.7%

MIssing 182 21.9% 115 21.0%Subject's birth weight

in grams 823 3,443 506 499 3,414 534

Subjects gestational

age in years 790 40.1 2.5 516 39.9 2.4

Multiple

pregnancies 21 2.5% 20 3.7%

Subject was breast

fed

Yes 406 48.9% 299 54.7%

No 254 30.6% 140 25.6%

Missing 171 20.6% 108 19.7%

Number of older

siblings

0 350 42.1% 262 47.9%

1 287 34.5% 163 29.8%

2+ 193 23.2% 119 21.8%

Missing 1 0.1% 3 0.5%

Sibling died 16 1.9% 15 2.7%

Parent(s) divorced,

seperated, or died

after subjects birth

Count and Percent 51 6.1% 32 5.9%

Family history of

schizophrenia

Yes 10 1.2% 4 0.7%

No 800 96.3% 536 98.0%

Missing 21 2.5% 7 1.3%

Family history of

post-traumatic

stress disorder

Yes 48 5.8% 31 5.7%

No 741 89.2% 500 91.4%

Missing 42 5.1% 16 2.9%

Familiy history of

bipolar disorder

Yes 66 7.9% 39 7.1%

No 727 87.5% 497 90.9%

Missing 38 4.6% 11 2.0%

Family history of

depression

Yes 323 38.9% 199 36.4%

No 458 55.1% 321 58.7%

Missing 50 6.0% 27 4.9%

Table 2


25


Review

Table 3





cited.

Outcome

Exposure

Category/Percentile % Yes Ratio Crude Risk Ratio 95% CI GEE* Risk Ratio 95%CI

Depression Any 19.8% 152/769 1.1 0.9-1.4 1.1 0.9-1.4

>=67th 19.0% 47/248 1.1 0.8-1.5 1.1 0.8-1.5

33rd <67th 20.8% 55/265 1.2 0.9-1.6 1.2 0.9-1.6

>0<33rd 19.5% 50/256 1.1 0.8-1.5 1.1 0.8-1.5

None 17.8% 92/518 Reference Reference Reference Reference

Bipolar Disorder Any 5.5% 36/653 1.7 0.9-3.2 1.8 0.9-3.5

>=67th 8.2% 18/219 2.6 1.3-5.1 2.7 1.3-5.6

33rd <67th 3.2% 7/217 1.0 0.4-2.5 1.1 0.4-2.7

>0<33rd 5.1% 11/217 1.6 0.7-3.5 1.6 0.7-3.7


Post Traumatic

Stress Disorder Any 7.1% 47/664 1.5 0.9-2.5 1.5 0.9-2.5

>=67th 8.2% 18/219 1.7 1.0-3.2 1.7 0.9-3.2

33rd <67th 6.7% 15/225 1.4 0.7-2.7 1.4 0.7-2.7

>0<33rd 6.4% 14/220 1.4 0.7-2.6 1.4 0.7-2.6


Schizophrenia Any 0.5% 3/620 2.1 0.2-20.0 2.1 0.2-20.0

>=67th 0.5% 1/202 ………….. ……….. ……….. ………..

33rd <67th 0.9% 2/212 …………… ………. ……….. ……….

>0<33rd 0.0% 0/206 ……….. ………. ……….. ……….


*Generalized estimating equations

Table 3


26


Review

Figures

Figure 1 Subjects by Mental Health Outcomes from Prenatal and Early Childhood

Exposure to Tetrachloroethylene and the Risk of Mental Illness by Aschengrau et al.

(2012).

Figure 2 Initial Selection from Selection, enrollment, and initial and final exposure

status of study subjects by Aschengrau et al. (2012).

27


Review

Figure 3 Final Selection from Selection, enrollment, and initial and final exposure

status of study subjects by Aschengrau et al. (2012).

Figure 4 Subjects by Status and Birth Year from Distribution of Selected Characteristics

of Subjects and Parents by PCE Exposure by Aschengrau et al. (2012).

28


Review

Figure 5 Subjects by Status, Gender, and Race from Distribution of Selected

Characteristics of Subjects and Parents by PCE Exposure by Aschengrau et al. (2012).

Figure 6 Subjects Available for Analysis from Distribution of Selected Characteristics of

Subjects and Parents by PCE Exposure by Aschengrau et al. (2012).

29


Review

Figure 7 Subjects by Education Status from Distribution of Selected Characteristics of


Figure 8 Subjects by Employment Status from Distribution of Selected Characteristics


30


Review

Figure 9 Subjects by Marital Status from Distribution of Selected Characteristics of


Figure 10 Subjects by History of Major Illness from Distribution of Selected


31


Review

Figure 11 Subjects by Solvent Exposure through Employment from Distribution of

Selected Characteristics of Subjects and Parents by PCE Exposure by Aschengrau et al.

(2012).

Figure 12 Subjects by Hobbies with Solvent Exposure from Distribution of Selected


32


Review

Figure 13 Age of Parents at Subjects Birth from Distribution of Selected Characteristics


Figure 14 Mother’s Education Level at Subjects Birth from Distribution of Selected


33


Review

Figure 15 Father’s Occupation at Subjects Birth from Distribution of Selected


Figure 16 Mother’s Prenatal Care at Subjects Birth from Distribution of Selected


34


Review

Figure 17 Mother’s Smoking Status at Subjects Birth from Distribution of Selected


Figure 18 Mother’s Alcohol Consumption at Subjects Birth from Distribution of Selected


35


Review

Figure 19 Mother’s Marijuana Usage at Subjects Birth from Distribution of Selected


Figure 20 Mother’s Medical Complications During Subjects Birth from Distribution of

Selected Characteristics of Subjects and Parents by PCE Exposure by Aschengrau et al.

(2012).

36


Review

Figure 21 Mother’s Occupational Solvent Exposure at Subjects Birth from Distribution

of Selected Characteristics of Subjects and Parents by PCE Exposure by Aschengrau et

al. (2012).

Figure 22 Subject’s Family History of Schizophrenia from Distribution of Selected


37


Review

Figure 23 Subject’s Family History of Post-Traumatic Stress Disorder from Distribution

of Selected Characteristics of Subjects and Parents by PCE Exposure by Aschengrau et

al. (2012).

Figure 24 Subject’s Family History of Bipolar Disorder from Distribution of Selected


38


Review

Figure 25 Subject’s Family History of Depression from Distribution of Selected


39


Review

End Notes

i

Persistent: Defined by Meriam Webster on-line (http://www.merriam-

webster.com/dictionary/persistent) as “existing for a long or longer than usual time or

continuously: degraded only slowly by the environment <persistent pesticides>.”

ii

For more information on Environmental Site Assessments see “Basic Elements

of Phase I and II Environmental Site Assessments” at

http://dnr.wi.gov/files/pdf/pubs/am/am465.pdf.

iii

The term “Acute” used in this context should be viewed using the standard

medical dictionary definition, as found at, http://medical-

dictionary.thefreedictionary.com/acute. Not to be confused with the term “Acute” toxic

hazardous waste which is reserved for “P” listed wastes. Tetrachloroethylene, when

disposed of as a hazardous waste is a “U” listed waste for toxicity under 40 CFR §

261.33, specifically U210, and not an acutely hazardous waste.

iv

For more information on generalized estimating equation see

https://onlinecourses.science.psu.edu/stat504/node/180.

v

For information on social stigmas of mental illness read “Understanding the

impact of stigma on people with mental illness” by Patrick W. Corrigan and Amy C.

Watson, World of Psychology, February 2002.

http://www.merriam-webster.com/dictionary/persistent

http://www.merriam-webster.com/dictionary/persistent

http://dnr.wi.gov/files/pdf/pubs/am/am465.pdf

http://medical-dictionary.thefreedictionary.com/acute

http://medical-dictionary.thefreedictionary.com/acute

https://www.law.cornell.edu/cfr/text/40/261.33

https://www.law.cornell.edu/cfr/text/40/261.33

https://onlinecourses.science.psu.edu/stat504/node/180

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1489832/

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1489832/

Documents

IH5426_Gutschmidt_J_EPI Paper Final Project