Increased prescriptions for irritable bowel syndrome after the 2018 Japan Floods: a longitudinal analysis based on the Japanese National Database of Health Insurance Claims and Specific Health Checkups

Large-scale natural disasters can cause not only mental stress [1, 2], but also gastrointestinal diseases due to unsanitary living conditions, food and water [3]. In fact, gastroduodenal ulcers [4, 5] and acute infectious diarrhea [6, 7] reportedly increase in the acute phase of natural disasters. In Japan, large-scale earthquakes such as the Great East Japan Earthquake and the Kumamoto Earthquake have occurred successively in recent years, and torrential rains have also occurred frequently around the world [8]. It is thus increasingly important for clinicians to understand the relationship between natural disasters and gastrointestinal diseases.

Irritable bowel syndrome (IBS) is known to be caused by mental stress, changes in gut bacterial flora, and inflammation of the bowel mucosa [9, 10]. For example, post-traumatic stress disorder (PTSD) suffered early in life reportedly increases the risk of future IBS [11]. Post-infectious IBS is one of the types of this syndrome, which develops after acute infectious diarrhea [12]. It has been reported that post-infectious IBS develops in approximately 10% of patients with acute infectious diarrhea [13]. Both mental stress and bowel infection often take place among victims of natural disasters. Therefore, we can infer that natural disasters can induce new onset of IBS or exacerbate existing IBS symptoms among disaster sufferers. Studies in the literature testing this hypothesis, however, is scarce.

IBS has a significant impact on the daily life of patients, and the proper and well-timed management of its symptoms significantly improves quality of life [14‐16]. IBS can result in anxiety, depression, and even suicide [13]. A higher risk of developing dementia among IBS patients over fifty years old has also been pointed out [17]. Thus, if IBS is triggered or worsen by natural disasters, the fact will be a scientific basis for clinicians to provide early and focused support for disaster victims who have had or are developing IBS symptoms.

Ramosetron, polycarbophil calcium, and mepenzolate bromide are approved solely for the treatment of IBS under the public insurance coverage in Japan; in other words, they cannot be prescribed for other diseases. These three drugs (IBS drugs) cannot be obtained over-the-counter in the market. Thus, a change of the prescription rate for these drugs is considered to reflect the change of apparent incidence, prevalence and severity of IBS among the population. Before a physician prescribes IBS drugs, a series of events must take place: the patient develops bowel symptoms, visits a physician, is diagnosed as IBS, and is judged as a case for applying one of the IBS drugs. Prescription is the final step in this process and indicates the existence of the underlying IBS diagnosis. We thus infer that if the number of prescriptions for IBS drugs among disaster victims is higher than non-victims, it is likely that the disaster causes IBS among the victims.

In terms of the scale of destruction, the 2018 Japan Floods was one of the largest natural disasters in Japan, with only the Great East Japan Earthquake causing more devastation in this century [18]. In the Japan Floods, 263 people died, 8 people went missing, 484 people were injured, 6783 houses completed destroyed, and 11,346 houses partially destroyed as of April 2019 [19]. The floods occurred across western Japan from June 28 to July 8, 2018, causing landslides and river flooding in many areas. Hiroshima, Okayama, and Ehime prefectures were particularly hit hard, with 238 deaths (approximately 90% of total deaths in the nation). The tremendous impacts, included death of family members, interruption of lifelines, and loss of housing forced victims to live in shelters, led to a substantial physical and psychological stress among victims.

This study aims to clarify the impact of the 2018 Japan Floods on the prescription rate of IBS drugs using a national universal health insurance database. This database contains all the prescriptions for IBS drugs and all disaster victims certified by their local governments. We conducted a comparison of the number of prescriptions between victims and non-victims and also between the pre- and post-disaster period. Based on the results, we will show the extent to which the disaster affected the incidence of newly treated IBS patients, how medication for IBS changed among patients treated by IBS drugs, and how long the effect lasted after the disaster.

The number of subjects included in this study was 5,287,888 out of the 6,176,299 registered in the NDB because subjects younger than 15 years of age were excluded. 32,499 people (0.61%) were categorized into the victim group. Table 1 shows baseline characteristics before the disaster. 50.5% of victims and 34.2% of non-victims were over 65 years old, and the proportion of women was slightly higher in the victim group. The largest number of victims in the study subjects were in Okayama prefecture, accounting for 46.5% of all victims (Additional file 1: Supplementary Table 1). Table 1 also shows non-users accounted for 99.1% of victims as compared with 99.4% of non-victims. Pre-disaster users also accounted for 0.88% of victims, which was higher than that of non-victims, suggesting that the prevalence of patients with IBS was higher in the victim than non-victim group even before the disaster. Victims among non-users were more likely to receive prescriptions before the disaster than non-victims, while the number was not a significant difference between victims and non-victims among pre-disaster users. Among pre-disaster users, the prescription rate of polycarbophil calcium was lower, and that of mepenzolate bromide was higher in the victim than in the non-victim group. There was no significant difference between the two groups in the prescribed days of IBS drugs before the disaster.

Table 2 shows the change of the type in IBS-related drugs among pre-disaster users. Probiotics were the most commonly prescribed in both groups before and after the disaster. Some groups of drugs were prescribed disproportionately more often in victims than in non-victims after the disaster; this is particularly the case for laxatives among all drug types.

Figure 1 shows the trend for IBS drug prescriptions among victims and non-victims before and after the disaster. The pre-disaster mean level difference between the victim and non-victim group was significant (0.09%; 95% CI 0.07–0.11; P < 0.001), and the difference in the mean baseline slope was also significant (− 0.005; 95% CI − 0.008 to − 0.001; P = 0.006). In the first month after the disaster (month 0), there appeared to be a significant increase in the prescription rate in the victim group: 0.05% (95% CI 0.02–0.09; P = 0.002). This was followed by a significant increase in the monthly trend for prescription rate (relative to the post-disaster trend for non-victim group): 0.01% per month (95% CI 0.004–0.015; P = 0.001) (Additional file 1: Supplementary Table 2).

Figure 2 shows the trend for prescriptions of IBS-related drugs among pre-disaster users. The pre-disaster mean level difference between the victim group and non-victim group was significant (4.48%; 95% CI 2.21–6.76; P < 0.001), but the difference in the mean baseline slope was not significant (0.02; 95% CI − 0.28 to 0.33; P = 0.87). In the first month of the disaster (month 0), there appeared to be no significant change in the prescription rate at − 0.39% (95% CI − 2.91 to 2.12; P = 0.75), followed by a monthly trend without a significant change in the prescription rate at 0.37% per month (95% CI − 0.06 to 0.79; P = 0.09), relative to the post-disaster trend for the non-victim group (Additional file 1: Supplementary Table 3). The occurrence of prescriptions for IBS-related drugs among victims gradually increased after the disaster, which was in contrast to the flat trend among non-victims.

Among non-users, the occurrence of a prescription for IBS drugs after the disaster was 0.71% for victims and was significantly higher than that for non-victims (0.35%, adjusted odds ratio 2.05; 95% CI 1.81–2.32) (Table 3). In subgroup analysis, a test for interaction terms was performed, and the interaction terms between gender and victim status was significant, while that between age group and victim status was not significant. After the disaster, the occurrence of prescriptions for IBS drugs was significantly higher in women (P for interaction = 0.01), and the heterogeneity among non-users was not found in terms of whether they were the elderly or not (P for interaction = 0.73).

In this study, we have shown that disaster victims were more likely to be prescribed IBS drugs after the disaster than non-victims. Particularly, among those who had no history of using IBS drugs before the disaster, victims were two-fold more likely to be prescribed IBS drugs than non-victims after the disaster. In other words, the new incidence of IBS supposedly increased, or the number of patients with untreated IBS who required drugs after the disaster increased among victims. Among pre-disaster IBS users, the trend for prescriptions of IBS-related drugs increased among victims after the disaster while it decreased among non-victims. Prescriptions of laxatives, in particular, increased after the disaster suggesting the increased incidence of constipation among victims with IBS.

This is the first study that demonstrated the impact of a large-scale natural disaster on prescriptions for IBS drugs. Another study investigated the incidence of IBS using the Rome III criteria, in middle-school students affected by the Wenchuan Earthquake in China. The incidence of IBS in the earthquake-affected area was reported to be the same as that in the non-affected area [21]. This earthquake study had several limitations: it was conducted as late as three years after the disaster, the sample size was small, there was no distinction between individuals affected directly by the disaster and those non-affected (i.e., the study was vulnerable to “ecological fallacy” in epidemiological terms). In order to overcome these limitations, we conducted this study using a large and reliable database that can separate victims from non-victims among the study population, which in turn has revealed a very different result from that of the previous study.

We assume three reasons for the increased incidence of IBS or the exacerbation of non-treated IBS. First, the disaster negatively impacted mental health among victims. IBS is often described as ‘brain-gut disorder’ because many patients with IBS also have psychiatric disorders and psychotropic drugs are effective for symptoms of IBS [22]. This functional relationship between the brain and gut is an important part of the pathophysiology of IBS [23]. When patients with IBS feel mentally stressed, gastrointestinal symptoms are aggravated [20]. In natural disasters, various mental stresses, due to loss of family or houses, directly affect victims, and the high level of the stress causes depression, insomnia and PTSD among the victims [1]. Subsequently, mentally stressed victims come to require IBS drugs for their bowel symptoms more often than non-victims. Second, the incidence of post-infectious IBS increases among victims because of acute infectious diarrhea caused by the disasters. Deterioration of sanitary conditions due to disasters can lead to an increase in the incidence of infectious diarrhea [6, 7]. The pathophysiology of IBS involves a change of gut microbiota, increased mucosal permeability, and mucosal inflammation; these are strongly related with post-infectious IBS [10]. In addition, it has been pointed out that mental stress, which is also part of the usual sequela of disasters, is a risk for developing post-infectious IBS [12]. Third, it is possible that IBS patients who received psychotherapy cannot continue the treatment after disasters because of reduced accessibility to professional psychotherapists. Psychotherapy is highly effective for symptom relief among IBS patients [24]. The interruption of the therapy potentially causes a worsening of IBS symptoms among victims.

The change in the prescription rate for IBS-related drugs among pre-disaster users may be caused by some food-related factors. Patients with IBS tend to form a habit of evading food that causes symptoms, and some of them even go further to undertake diet therapy such as antigen-eliminating diets based on IgG antibodies [25] and low FODMAP (fermentable oligosaccharides, disaccharides, monosaccharides and polyols) diets [26]. During a disaster, it is highly likely that they are unable to continue these diet patterns. For example, when living in a shelter, they are provided with ready-made or preserved foods, which would usually be avoided in their pre-disaster lives. These emergency foods can exacerbate symptoms of IBS. Constipation, which is caused by changes in food contents and reduced water consumption, would be particularly prevalent in such an emergency situation [3]. The finding in this study that laxatives are used more frequently than any other IBS-related drugs support this hypothesis.

It should be noted that, even before the disaster, victims were prescribed IBS drugs more often than non-victims. A reason for this would be that residents of flood-prone areas (e.g., riverbanks, lowlands, and mountain slopes), which usually are at a lower land price, are likely to be in a lower socioeconomic status. People with a low socioeconomic status are, in general, known to be at a higher risk for psychiatric disorders [27]. Since patients with psychiatric disorders are more likely to have comorbid IBS than those with normal mental health, people of a lower socioeconomic status would receive more IBS drug prescriptions. Thus, as shown in the results of this study, the disaster affected population with a higher IBS prevalence, which leads to an even further widening of IBS prevalence gap between victims and non-victims. In contrast, there may be accessibility bias between the victim and non-victim group. Victims more often received prescriptions for any medications before the disaster than non-victims, which indicates that they visited medical facilities more frequently. This means victims might be a group who had a better accessibility to medical facilities, and such a group of people would access facilities even more after the disaster. When interpreting our results, the escalation of pre-disaster high access among victims should be taken into account.

The frequency and intensity of natural disasters have increased in recent years, and thus physicians increasingly need to be familiar with disaster-related diseases. It is known that uncontrolled symptoms of IBS lead to decreased quality of life for patients with IBS [28]. Although symptoms of IBS may disappear spontaneously during the natural course of IBS [29, 30], early detection of new onset or exacerbation of IBS after disasters leads to better treatment and better quality of life for patients. Physicians should recognize that the JSGE guidelines recommend not to leave symptoms untreated and not to interrupt treatment [20]. Therefore, even in evacuation shelters, it is important to arrange space for victims to maintain privacy and reduce mental stress. In addition, we should consider to providing IBS patients with intestine-friendly and clean diets in order to prevent a worsening of IBS symptoms and acute infectious diarrhea. These measures should be included in policies to prepare for major disasters.

This study has some limitations. First, the diagnosis and the reasons for prescription are unknown because NDB data do not include such information. However, IBS drugs are not available over-the-counter, so people without a definite diagnosis by a physician cannot obtain IBS drugs. In other words, we can identify all patients with IBS who received treatment with IBS drugs. Second, ramosteron might be used for chemotherapy-induced nausea and vomiting or postoperative nausea and vomiting. Although it is used for this purpose in other countries, such use is not approved by Japanese public medical insurance. However, while such use of ramostrone may in fact exist in Japan, we did not exclude prescriptions for this drug according to this use. Third, the diagnostic criteria of IBS (e.g., the Rome IV criteria [31]) might have not been applied to all of our subjects taking IBS drugs. Thus, it is unclear whether the treated patients truly had criteria-met IBS. Fourth, all people affected by the disaster as well as the whole population who lived in the observed prefectures were not necessarily identified. In the NDB, only affected people who applied to the local government for victim certification and visited a medical institution were tagged with a victim code. Thus, we cannot identify affected people who did not apply or visit a medical institution. Furthermore, based on government statistics [32], the population of persons aged 15 and older of Hiroshima, Okayama, and Ehime prefecture as of October 2017 was estimated to be 2,461,000, 1,665,000, and 1,200,000, respectively, which added up to 5,326,000 people. The NDB used in this study identified 5,287,888 subjects, although these subjects included those who died and those who relocated during the observational period. By simple calculation, the number of subjects in this study covered approximately 99% (5,287,888/5,326,000) of the estimated population of the three prefectures, but this proportion might be too high. In one report [33], 26.5% of Japanese households had not visited any medical institution for two months, and in another report [34], 11.5% of households had not visited them for a year. According to these reports, although it is impossible to measure exactly the proportion of people who did not visit any medical institution during the observational period, we can estimate that the proportion was a few percent because the observational period was two years in this study, which was longer than that in the cited reports. In addition, even though there were such people, they would have been distributed in both the victim and non-victim groups in an unbiased manner. Thus, the impact of not being able to count the number of those who did not visit a medical institution would be negligible. Fifth, not all drugs used for IBS were analyzed in this study. Anxiolytics, anti-depressants, and herbal medicines (Kampo agents) are also used for this purpose in Japan. These drugs, however, are more often used for other diseases than IBS, and therefore we considered it inappropriate to include them in our analyses.

Finally, it is necessary to consider the possibility that the number of prescriptions of IBS drugs increased among victims because they did not have to pay for the medication themselves. The price elasticity of outpatient care in Japan’s health insurance system has been reported to be − 0.125 to − 0.076 [35]. The co-payment rate in the insurance system is generally 10% for people aged 75 and above, 20% for those aged 70–74, and 30% for those under 70. In this study, the mean co-payment rate for the victim group, based on age distribution, is 23.9%. When the co-payment was totally exempted, the increase in the amount of care demand was calculated as at 1.88–2.99%. Thus, the increase in demand for medical care due to the co-payment exemption would be minimal, and does not explain the substantial increase in prescriptions for IBS drugs.

The 2018 Japan Floods increased the number of prescriptions for IBS drugs among victims. The results suggest that natural disasters are a risk factor for IBS incidence. It is important for physicians who provide care for IBS patients to understand this risk and start comprehensive care including drug treatment in the early stages of a disaster.