Factors Associated with Infant Feeding Methods after the Nuclear Power Plant Accident in Fukushima: Data from the Pregnancy and Birth Survey for the Fiscal Year 2011 Fukushima Health Management Survey

The subjects of this study were 16,001 women who registered their pregnancies at municipal offices within or outside the Fukushima Prefecture between August 1, 2010 and July 31, 2011 and who underwent antenatal examination or gave birth on or after March 11, 2011. Self-administered questionnaires were mailed by post, and responses were received from 9321 of the 16,001 women (58.3 % response rate). Invalid responses and responses from women who had given birth before the disaster, resided outside the prefecture, had twins, or had suffered miscarriage or stillbirth were excluded. Thus, the responses from 8366 women were analyzed. This study was approved by the Ethical Review Committee of Fukushima Medical University (13051). The study objective was described on the cover of the questionnaire, and the return of the questionnaire was regarded as consent.

Statistical analyses were performed using IBM SPSS Statistics 21.0. Respondents who had used either formula feeding exclusively or a combination of breastfeeding and formula feeding were divided into two groups based on their reason for formula feeding, i.e., concern regarding radioactive contamination and other reasons (e.g. lack of breast milk). Various factors were compared between these two groups and the group that breastfed exclusively. Univariate analysis using the Chi square test was first carried out for the following factors: disaster-related factors (residential area, continuation of scheduled antenatal care, switched to another facility either within or outside the Fukushima Prefecture of their own accord), maternal factors (age, parity, depressive symptoms), obstetric factors (type of pregnancy, illness during pregnancy, delivery method), and child-related factors (sex of child, birth weight, congenital anomaly, neonatal asphyxia). Multiple logistic regression analysis was then used to investigate the association between disaster-related factors and formula feeding after adjusting for significant factors in the univariate analysis.

Subject characteristics are shown in Table 1. The mean subject age, mean gestational age, and mean birth weight were 30.1 ± 5.0 years, 39.0 ± 1.6 weeks, and 3030 ± 402 g, respectively. The mean elapsed time from delivery to questionnaire completion (baby age) and mean duration between disaster and delivery were 24.5 ± 13.6 weeks, and 23.4 ± 13.8 weeks, respectively. In addition, the minimum, 25th percentile, median, 75th percentile, and maximum values for subject age (15, 27, 30, 34, and 51, respectively), mean gestational age (22, 38, 39, 40, and 44, respectively), mean birth weight (402, 2788, 3034, 3280, and 4935, respectively), elapsed time from delivery to questionnaire completion (0, 14, 24, 35, and 92, respectively), and duration between disaster and delivery (0, 12, 23, 34, and 56, respectively) were determined.

Of the 8366 women surveyed, 2575 (30.9 %) had breastfed exclusively and 5765 (69.1 %) had both breastfed and formula-fed or formula-fed exclusively (Table 1). Among the latter group, the most common reason for formula feeding was insufficient breast milk (4166/5689; 73.2 %), followed by concern regarding radioactive contamination of breast milk (1154/5689; 20.3 %) and other reasons (1006/5689; 17.7 %).

The most common type of water used for formula preparation was bottled water, accounting for 75.5 % of cases (n = 4295; Table 2). Tap water was used significantly more often by women who formula-fed for reasons other than concern about radioactive contamination (p < 0.001), and bottled water was used significantly more often by those who formula-fed because of concern about radioactive contamination (p < 0.001).

Compared with women who breastfed exclusively (n = 2575), women who formula-fed because of concern regarding radioactive contamination (n = 1154) were significantly more likely to have the following characteristics (Table 3): resident of the evacuation area (p < 0.01), interrupted antenatal care (p < 0.001), depressive symptoms (p < 0.001), illness during pregnancy (p < 0.01), and cesarean section delivery (p < 0.001).

Compared with women who breastfed exclusively (n = 2575), women who formula-fed for reasons other than concern regarding radioactive contamination (n = 4535) were significantly more likely to be a resident of the evacuation area (p = 0.02), aged ≥ 35 years (p < 0.001), and primipara (p < 0.001). Moreover, the likelihood of formula feeding for those switching to another institution either within or outside the Fukushima Prefecture of their own accord was low (p < 0.001), but that for those having a cesarean section (p < 0.001) and a low-birth weight baby (p < 0.001) was high (Table 3).

Residence in the evacuation area [odds ratio (OR) = 1.28] and interrupted antenatal care (OR = 1.48) were significantly associated with formula feeding because of concern regarding radioactive contamination (Table 4). The ORs for residence in the evacuation area (OR = 1.40) and interrupted antenatal care (OR = 1.96) were even greater in a subanalysis of women whose only reason for formula feeding was concern regarding radioactive contamination (n = 622).

Residing in the evacuation area (OR = 1.20) and switching to a different medical institution (OR = 0.73) were associated with formula feeding for reasons other than radioactive contamination (Table 5).

The 2010 National Growth Survey on Preschool Children, a nationwide survey, found that the rate of exclusive breastfeeding was 50 % at all points between 1 and 5 months of age [12]. According to the 2000 National Growth Survey on Preschool Children, the breastfeeding rate after childbirth was 44.8 % at 1–2 months and 35.9 % at 4–5 months; the breastfeeding rate in Japan progressively increased in the 10 years thereafter [14]. Thus, the breastfeeding rate in our survey (30.9 %) was lower than the expected national figure. However, the breastfeeding rate in Fukushima Prefecture was reportedly lower than the national figure even before the disaster. In a 2007 survey by the Fukushima Bonyu no Kai (Fukushima Breastfeeding Association), the reported breastfeeding rate at age 4 months was 32.2 % [6].

From the survey data, two groups were created for analysis based on the duration between the disaster and the delivery. The group in which the period between the disaster and delivery was shorter showed a higher rate of formula feeding due to concern regarding the influence of radiation exposure on breast milk. Underlying the low breastfeeding rate in Fukushima were concerns regarding the transmission of radioactive substances to children through breast milk, which led mothers to formula-feed despite being able to breastfeed. The Infant and Child Feeding in Emergencies (IFE) Core Group coordinated by the Emergency Nutrition Network issued guidelines on “How to Ensure Appropriate Infant and Young Child Feeding in Emergencies” [9]. The IFE advised that the use of a nursing bottle and comforter should be actively discouraged in such disasters. Furthermore, the American Academy of Pediatrics (AAP) issued “Infant Feeding in Disasters and Emergencies” guidelines to provide guidance on radiation exposure [1]. The AAP reported that breastfeeding should not be stopped if an appropriate dose of iodine is administered at an appropriate time according to the advice of local health officials. Contamination of breast milk with iodine-131 was found even in women residing more than 200 km away from the Fukushima nuclear power plant [24]. However, according to the team that conducted the abovementioned government survey, the amounts of radioactive substances detected in the breast milk of seven women immediately after the disaster were extremely small, and continuation of breastfeeding would have no negative effect on their children’s health [18]. Opportunities must be created to educate pregnant women on the importance of breastfeeding and to motivate them to breastfeed [2]. In light of the nuclear accident, it is not only important to convey the advantages of breastfeeding but also to provide information on radioactivity in a user-friendly form to help women not to discontinue breastfeeding because of uncertainty over radioactive contamination. A clear national guideline is needed for nuclear disaster preparedness.

Our study results showed that more women who had resided in the evacuation area formula-fed than breastfed exclusively regardless of the reason for formula feeding. Stress affects the sympathetic nervous and adrenal systems, which diminish milk production and inhibit let-down [8]. Fatigue caused by evacuation, an unstable lifestyle in a new place, insufficient fluid intake, and poor breast milk secretion due to stress may have potentially contributed to the low breastfeeding rate. Soon after the disaster, an ample supply of formula was reportedly available. Some breastfeeding women supplemented breast milk with formula out of anxiety, while others chose to formula-feed because of hygiene-related concerns [19, 22]. In times of disaster, women may be more inclined to continue breastfeeding if carefully tailored support, in terms of providing adequate fluids, food, blankets, and other supplies, and ensuring privacy, is made available [4].

We found that a higher proportion of women whose antenatal care was interrupted formula-fed because of concerns regarding radioactive contamination. Previous studies have shown that a sense of coherence (the ability to counter stress for oneself) during pregnancy may be predictive of mental health during the puerperal period [20]. Furthermore, difficulties encountered during an unexpected event, such as a nuclear accident, may affect self-perception skills [11]. These studies suggest that particular attention should be paid to providing mental health care to women whose scheduled antenatal program is interrupted.

Contrary to antenatal care interruption, the breastfeeding rate was higher for those who switched to a different medical institution within or outside Fukushima Prefecture of their own accord. The higher breastfeeding rate among women who switched to a different medical institution may have been due to a greater sense of safety and self-efficacy. Previous studies have shown a high rate of exclusive and continued breastfeeding among women with high self-efficacy [15]. Increasing a women’s sense of self-efficacy and offering empowering advice may promote breastfeeding [2, 21]. Women whose medical care was interrupted may have been more vulnerable to the unexpected event and associated loss of control. Thus, efforts to provide such women with appropriate counseling and support to increase their sense of control should be undertaken.