4.1 Polychlorinated biphenyls, dibenzofurans and mortalities
A total of 1,803 Yucheng subjects with 48,751 person-years and 5,170
neighborhood referents with 141,774 were at risk from January 1, 1980 to the dates of death or December 31, 2008. A total of 295 Yucheng subjects and 757 neighborhood referents died during that time. Table 1 shows that there was a similar age and gender distribution for Yucheng subjects and neighborhood referents.
Table 2 shows results for all-cause and cause-specific mortality. The all-cause standardized mortality ratio was elevated for all Yucheng subjects and for males specifically. Elevations occurred among all Yucheng subjects for diseases of the circulatory system (SMR=1.3, 95% CI: 1.0–1.6). In diseases of the circulatory system, the SMRs for acute myocardial infarction (SMR=2.0, 95% CI: 1.0–3.4), other forms of heart disease (SMR=2.3, 95% CI: 1.4–3.5), cardiac dysrhythmias (SMR=5.8, 95% CI:
1.8–13.9), and late effects of cerebrovascular disease (SMR=2.9, 95% CI: 1.3–5.7) were increased. The SMR for diseases of the musculoskeletal system and connective tissue (SMR=6.4, 95% CI: 2.8–12.7) was much increased due to systemic lupus erythematosus mortality (1 male and 5 females in Yucheng subjects, 0 in neighborhood referents).
Among Yucheng males, the SMRs for diseases of the digestive system (SMR=1.9, 95% CI: 1.2–2.8), and injury and poisoning (SMR=1.5, 95% CI: 1.0–2.1) were
increased. In diseases of the digestive system, the SMR for chronic liver disease and cirrhosis (SMR=2.5, 95% CI: 1.5–3.9) was increased. Although the SMR for disease of the circulatory system was not increased, the SMRs for acute myocardial infarction
24
(SMR=3.3, 95% CI: 1.6–6.4) and other forms of heart disease (SMR=2.2, 95% CI: 1.2–
3.8) were increased. Although the SMR for all neoplasms was not increased, the SMRs for malignant neoplasm of stomach (SMR=3.5, 95% CI: 1.5–7.0) and malignant neoplasm of lymphatic and haematopoietic tissue (SMR=3.0, 95% CI: 1.1–6.6) were increased. The SMR for diabetes mellitus was decreased (SMR=0.3, 95% CI 0.1–0.9).
Among Yucheng females, the SMRs for diseases of the circulatory system (SMR=1.5, 95% CI: 1.0–2.1) and musculoskeletal system and connective tissue
(SMR=16.5, 95% CI: 6.7–34.3) were increased. In disease of the circulatory system, the SMRs for other forms of heart disease (SMR=2.4, 95% CI: 1.2–4.5), cardiac
dysrhythmias (SMR=9.6, 95% CI: 2.4–26.0), and late effects of cerebrovascular disease (SMR=5.4, 95% CI: 1.7–13.1) were increased. In disease of musculoskeletal system and connective tissue, the relative mortality from systemic lupus erythematosus was very high (5 in Yucheng females, 0 in neighborhood referents).
Compared with our previous study (Tsai, et al., 2007), new findings here are increased mortality from all causes, malignant neoplasms, and diseases of the circulatory system (Table 3).
25
4.2 Mortalities among Yucheng and Yusho victims
Table 4 summarizes the characteristics of Yucheng and Yusho cohorts. A total of
1,803 Yucheng subjects (Male, N=830; Female, N=973) with 48,751 person-years of
follow-up and 1,664 Yusho subjects with 50,773 person-years are included. A total of
295 Yucheng subjects and 441 Yusho subjects died during the follow-up period.
Table 5 showed the cause specific mortalities from the individual studies, as well as
the pooled mortalities. An increase in all-cause mortality was found in pooled subjects
(pooled SMR=1.1, 95% CI: 1.1–1.2, I2=28.6%), notably among pooled males (pooled
SMR=1.2, 95% CI: 1.1–1.3, I2=0.0%). Pooled male and pooled female subjects had
different findings in regard to all cancer mortality. The SMR for all cancer was
increased only in pooled males (pooled SMR=1.3, 95% CI: 1.1–1.6, I2=0.0%). Despite
that liver cancer was not elevated in pooled subjects (pooled SMR=1.5, 95% CI: 0.9–
2.7, I2=50.1%), significant elevation was found in pooled females (pooled SMR=2.0,
95% CI: 1.1–3.6, I2=0.0%). Elevation was found for lung cancer (pooled SMR=1.5,
95% CI: 1.1–2.1, I2=0.0%) among pooled subjects, notably among pooled males
(pooled SMR=1.7, 95% CI: 1.2–2.3, I2=0.0%).
Elevations occurred in pooled subjects for heart disease (pooled SMR=1.3, 95% CI:
1.0–1.7, I2=43.4%). Increased hepatic disease mortality was found in pooled subjects
26
(pooled SMR=1.5, 95% CI: 1.0–2.4, I2=27.7%), notably among pooled males (pooled
SMR=1.9, 95% CI: 1.3–2.8, I2=0.0%).
27
4.3 Gestational exposure to polychlorinated biphenyls and hearing loss
Among the 184 Yucheng children invited, 86 agreed to participate in this
examination. Among the 184 referent children, 97 agreed to participate. The Yucheng
and referent children were of similar age, gender, body mass index, total cholesterol,
and triglyceride (Table 6). Among Yucheng children, non-participants had average age
of 21.3 ± 3.8, 51.0% males. Among referents, non-participants had average age of 21.3
± 3.8, 51.7% males. These were not different from participants (data not shown).
For pure tone auditory at frequencies from 250 Hz to 8000 Hz, hearing threshold of
> 20 dB was more frequently observed at 250 Hz, 500 Hz, and 2000 Hz among Yucheng
children as compared to the referents. Estimated maternal concentrations of PCBs and
PCDFs at the time of birth are shown in table 7.
Table 8 showed results of logistic regression using elevated pure tone auditory
thresholds (>20 vs. <=20) at different frequencies as dependent variables, and exposure
status (Yucheng vs. referent), age, gender, body mass index, total cholesterol, and
triglyceride as independent variables. Yucheng children were at higher risk of elevated
hearing threshold in the right ear at frequencies 250 Hz, 500 Hz, and 2000 Hz, as
compared to the referents.
Table 9 showed linear regression using log-transformed pure tone auditory
28
thresholds as dependent variable and log-transformed maternal serum concentrations of
PCDFs and marker-PCBs as independent variable. Maternal concentrations of
2,3,4,7,8-pnCDF at the time of delivery were found associated with pure tone auditory thresholds
at frequency 250 Hz, 500 Hz, 1000 Hz, and average threshold level of right ear, and 500
Hz, 4000 Hz, and average threshold level of the left ear, after adjusting for age, gender,
body mass index, total cholesterol, and triglyceride.
Maternal concentrations of 1,2,3,4,7,8-hxCDF were associated with pure tone
auditory thresholds at frequency 4000 Hz of the left ear. There was no association found
between maternal marker-PCB concentrations and pure tone auditory thresholds.
Table 10 showed linear regression using log-transformed DPOAEs as dependent
variable and maternal serum concentrations at pregnancy of log-transformed maternal
PCDFs and marker-PCBs concentrations (at birth) as independent variable. Maternal
concentrations of 2,3,4,7,8-pnCDF at delivery were found negatively associated with
DPOAE amplitudes at 1500 Hz, 2000 Hz, average amplitude of right ear, and average
amplitude of left ear. The results were unchanged when adding 1,2,3,4,7,8-hxCDF or
the sum of marker-PCBs in the model of regression analysis (data not shown).
29