Psychological Distress in Workers at the Fukushima Nuclear Power Plants FREE

To the Editor: A magnitude 9.0 earthquake and tsunami on March 11, 2011, triggered plant explosions and a nuclear meltdown at the Fukushima Daiichi nuclear power plant. The nearby Daini nuclear power plant also experienced damage but remained intact. Studies after the nuclear disaster at Chernobyl¹ suggest nuclear power plant workers are at risk for general psychological distress, including posttraumatic stress response (PTSR). We examined the psychological status of Fukushima workers 2 to 3 months after the disaster.

Following approval by the ethics committees of Ehime University and National Defense Medical College, we recruited all full-time nuclear power plant workers from the Daiichi (n = 1053) and Daini (n = 707) plants in May and June 2011. Written informed consent was obtained.

Using a self-report questionnaire, we assessed sociodemographic characteristics and disaster-related experiences (Table 1; coded dichotomously as “yes” or “no”), including discrimination/slurs (sabetsu/chuushou) because the electric company was criticized for their disaster response and the workers have been targets of discrimination.² General psychological distress was evaluated using the K6 scale (Japanese version),³ including items on feeling nervous, hopeless, restless/fidgety, depressed, everything was an effort, and worthless in the last 30 days. Scores ranged from 0 to 24, with 13 or higher indicating high distress.³ PTSR was assessed by the Japanese version of the Impact of Event Scale Revised (IES-R-J), a 22-item scale including PTSR domains of intrusion, avoidance/numbing, and hyperarousal.⁴ Scores ranged from 0 to 88, with 25 or higher indicating high PTSR.⁴ Cronbach α was high for K6 (0.88) and IES-R-J (0.95).

Two-tailed χ² tests were performed to evaluate the difference in proportions. Significant independent variables from bivariate analysis were considered potential factors of high general psychological distress and PTSR, and were entered in the multivariable logistic regression model (forced entry method). SAS version 9.2 (SAS Institute) was used. A 2-sided P < .05 was used to indicate significance.

Of 1760 eligible workers, 1495 (85%) participated (Daiichi: n = 885 [84%]; Daini: n = 610 [86%]). Compared with Daini workers, Daiichi workers were more often exposed to disaster-related stressors (Table 1). Experiencing discrimination or slurs was not statistically significantly different between groups (14% vs 11%, P = .08).

Daiichi workers had significantly higher rates of psychological distress (n = 412; 47%; 95% CI, 43%-50%; vs n = 226; 37%; 95% CI, 33%-41%; P < .001) and PTSR (n = 261; 30%; 95% CI, 27%-33%; vs n = 117; 19%; 95% CI, 16%-22%; P < .001) (Table 1). For both groups, discrimination or slurs were associated with high psychological distress (Daiichi: adjusted odds ratio [AOR], 2.06; 95% CI, 1.34-3.16; vs Daini: AOR, 2.90; 95% CI, 1.63-5.17) and high PTSR (Daiichi: AOR, 2.17; 95% CI, 1.43-3.30; vs Daini: AOR, 2.70; 95% CI, 1.47-4.96) (Table 2). Other significant associations in both groups included tsunami evacuation and major property loss with psychological distress and preexisting illness and major property loss with PTSR.

We found that general psychological distress and PTSR were common in nuclear plant workers 2 to 3 months after the disaster. The prevalence was higher than in other studies (12.5% with severe or very severe psychological impairment in a review of 24 studies),⁵ possibly due to the complexity of their experience. Higher rates were found among workers of Daiichi than Daini, which is concordant with their higher exposure to disaster-related stressors.

This is the first study to our knowledge to explore discrimination as a factor in postdisaster mental health. Experiencing discrimination was associated with both general psychological distress and PTSR. A similar phenomenon was observed in Vietnam War veterans; along with combat exposure, insufficient societal support and societal rejection upon homecoming were associated with posttraumatic stress disorder.⁶

Several limitations warrant discussion. Our report was cross-sectional, with neither baseline measures nor long-term outcomes. The responses were self-reported and no comparison group was available. We had no information on specific previous physical/mental illness; educational, marital, or socioeconomic status; or precise irradiation exposure, although irradiation symptoms were not reported.