Burden and Management of Noncommunicable Diseases After Earthquakes and Tsunamis

Abstract

This integrative review examines extant literature assessing the burden and management of noncommunicable diseases 6 months or more after earthquakes and tsunamis. We conducted an integrative review to identify and characterize the strength of published studies about noncommunicable disease–specific outcomes and interventions at least 6 months after an earthquake and/or tsunami. We included disasters that occurred from 2004 to 2016. We focused primarily on the World Health Organization noncommunicable disease designations to define chronic disease, but we also included chronic renal disease, risk factors for noncommunicable diseases, and other chronic diseases or symptoms. After removing duplicates, our search yielded 6,188 articles. Twenty-five articles met our inclusion criteria, some discussing multiple noncommunicable diseases. Results demonstrate that existing medical conditions may worsen and subsequently improve, new diseases may develop, and risk factors, such as weight and cholesterol levels, may increase for several years after an earthquake and/or tsunami. We make 3 recommendations for practitioners and researchers: (1) plan for noncommunicable disease management further into the recovery period of disaster; (2) increase research on the burden of noncommunicable diseases, the treatment modalities employed, resulting population-level outcomes in the postdisaster setting, and existing models to improve stakeholder coordination and action regarding noncommunicable diseases after disasters; and (3) coordinate with preexisting provision networks, especially primary care.

This integrative review examines extant literature assessing the burden and management of noncommunicable diseases 6 months or more after earthquakes and tsunamis that occurred from 2004 to 2016. Results demonstrate that existing medical conditions may worsen and subsequently improve, new diseases may develop, and risk factors, such as weight and cholesterol levels, may increase for several years after an earthquake and=or tsunami.

Disasters, both natural and man-made, have a wide range of health and human impacts.^1-4 Many natural disasters have acute destructive potential, leading to a significant burden of wounds and injuries.⁵ Especially in developing countries, natural disasters can lead to severe infrastructure and resource disruption, as well as displacement, contributing to communicable disease outbreaks.^6,7 Increased environmental exposure from displaced, or in the case of nuclear disasters, primarily released, chemicals can cause acute and chronic health effects.⁸ The psychological impact of disasters—due to experiencing the event itself, loss of loved ones, or media exposure—is often present and can manifest acutely, chronically, or both.^9-12 As a result of these health effects, disasters can have significant impacts on quality of life.¹³

The impact of natural disasters on chronic disease has been increasingly discussed in the literature.^14-17 Medication shortages, interruptions in care, displacement, and destruction of critical infrastructure have been identified as key barriers to care in the postdisaster phase.^18-23 Regardless of the progress in this field, a recent review examining recommendations for chronic illness management in disasters concluded that more evidence-based recommendations are needed, and the shortage of evidence on noncommunicable disease interventions in humanitarian settings has also recently been highlighted.^24,25 Existing approaches to address these issues focus primarily on prevention of acute exacerbations of existing noncommunicable diseases and interruptions in treatment, while recognizing that, ideally, these efforts would synergize with existing efforts to improve noncommunicable disease care outside of disaster settings.^15,26,27

Much of the extant research examines noncommunicable disease–specific outcomes and interventions immediately or soon after a disaster.^28-34 For example, acute health impacts of earthquakes 3 months postdisaster have been characterized elsewhere and focus primarily on traumatic injury and its sequelae.² However, it was recently suggested that acute and long-term complications related to noncommunicable diseases postdisaster need to be better characterized in terms of their morbidity and mortality.^14,24

In this integrative review, we address these gaps by examining the research literature on noncommunicable disease–specific outcomes and interventions in the more protracted phase after a disaster. Because noncommunicable diseases are chronic illnesses that develop, stabilize, and evolve over longer periods, we focus on outcomes that manifested 6 months or more after a disaster. We focus on earthquakes and tsunamis because of their tremendous potential for human impact and physiological and emotional stress, all of which are thought to mediate worsening outcomes for those with existing disease and the development of new disease.^2,35-37 Although they manifest differently, both tsunamis and earthquakes are caused by the abrupt shifting of tectonic plates. Studies have examined how individuals and health systems might respond to noncommunicable diseases in a disaster setting, but to our knowledge this is the first effort to rigorously review both the content and quality of literature and characterize how noncommunicable diseases develop or deteriorate in the non-acute phase after a specific type of disaster. In addressing this gap, we aim to contribute to the formation of specific, evidence-based plans and recommendations for noncommunicable disease management in postdisaster settings beyond the immediate disaster phase.

Methods

Prior to conducting this integrative review, we identified reviews focused on: (1) existing disaster preparedness recommendations and guidelines for chronically ill patients; (2) noncommunicable disease outcomes after cyclone-, storm-, and flood-related disasters; and (3) select impacts of natural disasters on cardiac, renal, and diabetic patients. This review proceeded given that a different type of disaster was being considered, as well as the fact that one of the identified reviews, and several other publications, have called for more evidence to inform postdisaster, noncommunicable disease–related recommendations.^{14,24,25,38,39}

We adapted our methodology from that of several published integrative reviews that focused on different aspects of emergency preparedness.^40,41 We sought to identify and characterize the strength of published studies about noncommunicable disease–specific outcomes and interventions at least 6 months following an earthquake or tsunami.

Search Strategy

We searched the following health-oriented databases: PubMed, EMBase, CINAHL, SCOPUS, and Web of Science. We initially developed our search terms using a priori knowledge and experience with prior research in this area.³⁶ To refine our search terms, we consulted an informationist.^*

The search was performed on February 15, 2017. The date range for the search was January 1, 1970, to February 15, 2017.

Inclusion Criteria

Articles were included if they were original peer-reviewed quantitative or qualitative studies published in English that considered diagnosis, management, and/or outcomes of noncommunicable diseases at least 6 months after an earthquake or tsunami. Abstracts, conference proceedings, and grey literature in the form of reports or anecdotes were not included. We defined noncommunicable diseases in accordance with the World Health Organization (WHO) as cardiovascular disease, respiratory disease, cancer, and diabetes, although we also recognized that noncommunicable diseases refer broadly to nontransmissible conditions.⁴² We included various forms of renal disease, given that its 2 most common causes are hypertension and diabetes, both considered noncommunicable diseases under the WHO designation.⁴³

Compared to communicable disease, a distinction of noncommunicable diseases is the inability to biologically transmit these illnesses, although data have shown that familial patterns of behavior can contribute to risk, incidence, and prevalence of noncommunicable disease risk factors, such as smoking and dietary patterns.^44-46 Therefore, studies addressing clear and known risk factors for noncommunicable diseases (eg, smoking, body weight) were included, as were studies that highlighted chronic, nonspecific, nontransmissible symptoms present at least 6 months postdisaster. Although high body mass index (BMI) and smoking are considered risk factors, they are also medical conditions (eg, obesity, addiction) with a long history of causal association with several of the WHO's noncommunicable disease designations.^47-49

Because much of the literature in this area is anecdotal, commentary, or experience-driven, or based on small studies or case reports, we adopted a 50-subject minimum for study inclusion, previously used in literature addressing a similar topic.^2,50-57 If a notable finding existed in a study that addressed noncommunicable diseases and other outcomes (eg, those not involving noncommunicable diseases), only the relevant information was included in this review.

Exclusion Criteria

Our primary exclusion criterion was literature focused on evaluating mental health outcomes postdisaster. In contrast, we included literature in which psychiatric, psychosocial, or psychological variables were assessed as predictors of or having relationships with other noncommunicable diseases being assessed as outcomes. During our title review phase, we excluded studies in which the primary focus was the short- or long-term burden of wounds, injuries, and surgical conditions, as this has recently been explored.²

During our abstract review phase, we applied the following additional exclusions: literature evaluating earthquakes and tsunamis that occurred before 2004; addressing nonspecific quality of life measures; chronic pain; long-term injury-related disability and rehabilitation burden, outcomes, and interventions; obstetric and/or gynecological outcomes; and nutritional status or interventions. We selected 2004 as the cutoff event year because of the concentration of literature around major disasters that occurred in and after that year, including the 2004 Indian Ocean Earthquake, the 2010 Haiti Earthquake, and the 2011 Great East Japan Earthquake. The various content areas were excluded either because existing reviews were identified or the area was determined to be a separate and distinct field of study and research beyond the scope of noncommunicable diseases.⁵⁸ We excluded studies related to the Great East Japan Earthquake that primarily examined the effects of the subsequent radiologic disaster (eg, radioactive environmental exposure or impact of evacuation associated primarily with residence in the evacuation zone) rather than evacuation or displacement associated with the earthquake or tsunami itself.

During abstract review, we selected 6 months as the postdisaster outcome threshold, given previous research examining the impact of earthquakes within 3 months and given our previous rationale that assessing impact on this time horizon would more accurately characterize longer-term noncommunicable disease outcomes once the burden of wounds, injuries, communicable disease, and acute exacerbations of chronic illness had lessened.² Given the paucity of this literature, if noncommunicable disease outcomes were assessed both prior to and after 6 months postdisaster in the same study, articles were included, and our synthesis and discussion of them focused primarily on results occurring at least 6 months postdisaster. If studies collected data up to, but not past, 6 months, data from the 6-month mark were included.

Article Review Protocol

Our initial search yielded 9,494 results. After removing duplicates, 6,188 results remained (Figure 1). All titles were screened for relevance by one co-author (AS) and concurrently reviewed by a second co-author (LR or MG). Differences in recommendation for title inclusion were reconciled via discussion and consensus among co-authors. After title review, 1,134 articles remained for abstract review. Two co-authors (AS and LR) reviewed all abstracts, with differences in recommendation for abstract inclusion reconciled via discussion and consensus among co-authors. After abstract review, 75 articles remained for full text review. After full text review, 25 articles remained. The most common reasons for exclusion after full text review were: (1) on further examination, the item was an abstract or poster (n = 21); (2) the article examined the impact of the nuclear disaster resulting from the Great East Japan Earthquake rather than the earthquake or tsunami itself (n = 10); and (3) data were collected less than 6 months postdisaster (n = 8).

Figure 1.

Review Protocol

Each of the 25 final articles was evaluated with an adapted version of criteria used in previous disaster preparedness research as well as other fields (see Figure 1).^40,41,59,60 Drawing on these criteria, our assessment of methodological rigor focused on sampling, reliability, validity, and bias. We assessed pretesting as part of our measure of reliability and evaluation of instrument development as part of our measure of validity.

Although qualitative data are not often included in such reviews, they were included here to deepen the nature of our results by including subjectively experienced and/or non–formally diagnosed noncommunicable diseases (eg, chronic physical symptoms).

Results

A total of 25 peer-reviewed articles met inclusion criteria for this review. The distribution of disaster events was: 2011 Great East Japan Earthquake and tsunami (n = 13); 2008 Sichuan Earthquake (n = 3); 2004 Mid Niigata Prefecture Earthquake (n = 3); 2010-11 Christchurch earthquakes (n = 2); 2004 Indian Ocean tsunami (n = 2); other events (n = 2). The articles were published between 2006 and 2016. Data were collected from 6 months to 6 years postdisaster (Table 1).

Table 1.

Overview of Studies by Disaster, Purpose, Methods, Sample and Population Characteristics, Timeframe, and Key Findings

Citation (publication year)	Disaster (event year)	Purpose	Methods	Sample	Population	Timeframe
Earthquake events only
Erskine et al (2013)	Canterbury Earthquakes, New Zealand (2010 & 2011)	Assess changes in smoking status and level of tobacco consumption postdisaster	Semi-structured interviews	1,001	Residents of Christchurch in high-flow pedestrian areas	15 months after beginning of earthquake activity in Canterbury
Key findings: 24% of ex-smokers resumed smoking 15 months after earthquake, 34.1% of current smokers had increased consumption levels compared to pre-earthquake levels, and of those who acknowledged reasons for the increase, 76.8% cited earthquake as contributory. 31.5% of those smoking before earthquake decreased consumption, citing various reasons including health concerns and peer/family pressure.
Huang et al (2016)	Sichuan Earthquake, China (2008)	Assess impact of disaster on number, cost of hospitalizations for ischemic heart disease	Retrospective chart review	19,083	Individuals hospitalized in Deyang, a city heavily affected by the disaster	3 years predisaster to 6 years postdisaster (2005-2014)
Key findings: During the 5 years postdisaster (2008-2012), rates of ischemic heart disease–related hospitalization continued to increase in both hard hit and extremely hard hit areas (RR 1.32 in 2008, 1.93 in 2009, 2.02 in 2010, 2.19 in 2011, 1.26 in 2012), although decreased in the 2 years thereafter.^* Cost of hospitalizations increased but was not statistically significant.
Kamoi et al (2006)	Mid Niigata Prefecture Earthquake, Japan (2004)	Determine impact of disaster on HbA1c levels and complica-tions in type 1 diabetic patients	Monthly physical exam and laboratory examinations, 1-time self-report questionnaire (plus predisaster data)	65	Patients visiting diabetes clinic at Nagaoka Red Cross Hospital	Monthly for 12 months postdisaster
Key findings: Twelve months postevent, HbA1c levels returned to pre-earthquake levels for patients treated with multiple daily insulin injections, although remained elevated (<0.5%) for patients treated with continuous subcutaneous insulin injections.^** In this study, many patients switched from disposable insulin pens requiring refills to pre-filled injection pens, and complications postdisaster were minimal.
Kamoi et al (2006)	Mid Niigata Prefecture Earthquake, Japan (2004)	Assess impact of disaster on morning home blood pressure measurement, glycemic control, and diabetes complications	Physical and laboratory exams every 1-3 months post-disaster, 1-time self-report questionnaire (plus pre-disaster data)	222	Patients visiting diabetes clinic at Nagaoka Red Cross Hospital	Up to 6 months postdisaster
Key findings: Six months postdisaster, 71% of those who previously took home blood pressure measurements had resumed. At that time, those who did not measure their morning home blood pressure had higher HbA1c levels compared to predisaster levels.^* Those who did measure their home morning blood pressure had higher urinary albumin excretion rates compared to predisaster.^* There were no differences in rate of complications between those who did and did not measure their blood pressure 6 months postdisaster.
Khashram et al (2016)	Christchurch Earthquake, New Zealand (2011)	Assess whether patients received timely carotid endarterectomy (CEA) consistent with established hospital protocol for internal carotid artery stenosis postdisaster	Retrospective chart review	62	All symptomatic patients undergoing carotid endarterectomy in Christchurch Hospital	Up to 17 months postdisaster
Key findings: Postdisaster in 2011, CEA was performed within recommended 2 weeks for 76% of patients. Compared to 2008 when a protocol for referral to CEA was established, average wait time to CEA decreased by 23 days. The largest gap was from presentation to carotid imaging, primarily among outpatients.
Nakagawa et al (2009)	Mid Niigata Prefecture Earthquake, Japan (2004)	Assess whether death from acute myocardial infarction increased in disaster-affected areas	Examined death records	2,448,025	Disaster areas (12 affected municipalities within Chuetsu region) and control areas (unaffected municipalities within Niigata Prefecture)	5 years predisaster and up to 3 years postdisaster
Key findings: Compared to 5 years predisaster, mortality from acute myocardial infarction significantly increased (14%) 3 years postdisaster in disaster areas compared to nondisaster areas.^*
Omote et al (2013)	Noto Peninsula Earthquake, Japan (2007)	Assess postdisaster changes in hemoglobin, hematocrit, and RBC associated with degree of property damage	Annual health screenings	5,563	Residents of Wajima, an affected city in Japan	Up to 12 months postdisaster
Key findings: RBCs, Hb, and Ht decreased postdisaster generally for males and females.^* No relationship was elucidated between property damage and outcomes.
Sun et al (2013)	Sichuan Earthquake, China (2008)	Assess risk factors and prevalence of hypertension postdisaster	Physician-reported surveys and physical examination	3,230	Individuals >20 years old in temporary, postdisaster shelter for more than 1 year from 2 regions in China	9 to 12 months postdisaster
Key findings: Age, family history of hypertension, diabetes, BMI, waist-hip ratio, sleep quality, and blood glucose were risk factors for development of hypertension after earthquake (overall prevalence 24.08%).^* The experience of mental stress, lower income, and housing or property damage was not different between hypertensive and nonhypertensive groups.
Wen et al (2012)	Sichuan Earthquake, China (2008)	Assess physical diseases and psychological impairment among survivors 3 years postdisaster	Interviewer-administered questionnaires	2,525	Individuals >7 years old in households in earthquake-stricken rural areas—both hard-hit and less-hit counties (3 in each category)	3 years postdisaster
Key findings: Prevalence of chronic illness was higher in hard-hit areas (39.2%) compared to less-hit areas, as was 2-week prevalence of chronic disease morbidity (24.9%).^*
Tsunami events only
Keskinen-Rosenqvist et al (2010)	Indian Ocean Tsunami (2004)	Assess relationship between various psychological exposure variables and physical symptoms	Mailed questionnaire	1,505	Residents of Stockholm 16 years and older present during 2004 Indian Ocean Tsunami who then returned home	14 months postdisaster
Key findings: Several groups, including those with various combinations of life threatening exposure, bereavement, and severe injury, reported increased physical symptoms in postdisaster period (OR 1.5-6.8).
Turner et al (2009)	Indian Ocean Tsunami (2004)	To compare postdisaster health between individuals living in housing projects versus transitional camps	Interviewer-led questionnaire	303	Individuals over age 18 living in 7 transitional camps and 5 permanent housing projects across 3 districts	26-27 months postdisaster
Key findings: Two years postdisaster, living in transitional housing (vs housing project) was a risk factor for poorer self-rated health among men and women.^* More specifically, it was also a risk factor for cough, stomach ache, headache, general aches and pains, and feeling unwell.^**
Combined earthquake and tsunami events
Furusawa et al (2011)	Solomon Islands Earthquake and Tsunami (2007)	Assess various communicable and noncommunicable health outcomes several years postdisaster	Self-report questionnaire and health check-up	160 households	4 villages across 3 islands in the Solomon Islands	27 months postdisaster
Key findings: Prevalence of obesity, hypertension, and/or diabetes varied across villages but was as high as 32.5%, 45.8%, and 8.9%.
Hagiwara et al (2016)	Great East Japan Earthquake and Tsunami (2011)	Assess low back pain postdisaster based on changes in living environment and working status	Self-report questionnaire	986	Residents >18 years old living in severely damaged coastal regions in Miyagi Prefecture	7-11 months and 22-23 months postdisaster
Key findings: Lower income and having same occupation after disaster were associated with low back pain several months postdisaster for those under 65 years of age and those above 65 years of age, respectively (OR 1.84, OR 2.05).^**
Hasegawa et al (2015)	Great East Japan Earthquake and Tsunami (2011)	Assess impact of disaster on burden of neuro-otologic diseases compared to non-neuro-otologic diseases	Retrospective chart review	18,167	New patients at Soma General Hospital	From 1 year before to 3 years postdisaster
Key findings: Presentation of several neuro-otologic diseases increased initially postdisaster but plateaued 2 years postdisaster and began to decrease 3 years after.^** Other otolaryngological diseases did not change in frequency, nor did overall number of patients seen. More diseases were complicated by comorbid mental health conditions compared to predisaster.
Kikuya et al (2015)	Great East Japan Earthquake and Tsunami (2011)	Assess asthma and eczema symptoms in children after disaster	Parent-administered questionnaire	1,277	Children (2nd/4th/6th/8th grade students) in 3 municipalities in Miyagi Prefecture	8 months postdisaster
Key findings: Eight months postdisaster, 11.4% and 3.5% of children had current and severe wheezing symptoms, respectively; 15.6% and 1.5% had current and severe eczema symptoms. Rates of eczema were higher than the national average.
Konno et al (2013)	Great East Japan Earthquake and Tsunami (2011)	Assess differences in postdisaster blood pressure between public and nonpublic employees	Annual health check-ups in 2010 and 2011	240 public employees, 1,776 general population	General population in Watari, a city heavily affected by disaster	4 to 8 months postdisaster
Key findings: Systolic and diastolic blood pressure increased (11.3 mm Hg and 7.8 mm Hg, respectively) among public employees, a higher increase than the general population 4-8 months postdisaster.^*
Leppold et al (2016)	Great East Japan Earthquake and Tsunami (2011)	Assess relationship between socio-demographic factors and glycemic control in patients with diabetes postdisaster	Data extracted from patient charts and hospital laboratory records	404, but 284 analyzed after excluding those lost to follow-up	Residents over age 20 receiving treatment for type 1 or type 2 diabetes at public hospital in Minamisoma City before disaster	1 year predisaster and 12 months postdisaster
Key findings: Proportion of patients with poor glycemic control (HbA1c >7.0%) increased to 41.4% from 31.9% from 2010 to 2012.^ 66.5% of patients experienced deterioration of HbA1c (increased by >0.5%) during at least 1 postdisaster assessment point. Increasing age (OR 0.95), female gender (OR 0.50), and rural residence (OR 0.34) were “protective” against deteriorations in glycemic control.^
Miyashita et al (2015)	Great East Japan Earthquake and Tsunami (2011)	Assess wheeze and eczema symptoms of children in disaster-affected areas	Parent-administered questionnaire	7,155	Children in 13 municipalities in Miyagi Prefecture	Up to 3 years postdisaster
Key findings: Younger age, history of hospitalization, and worse SDQ score were associated with all allergic symptoms with few exceptions.^ Attending school in coastal municipality and tsunami experience were associated with current eczema and severe eczema, respectively (OR 1.18, OR 1.74).^ In coastal areas, wheezing was higher among children whose homes were rebuilt in same area, and eczema was more prevalent among children living in rental housing without formal leasing system for disaster-affected populations.^ In inland areas, wheezing was more prevalent among children living in rental housing without formal leasing system for disaster-affected populations or living with friends and family.^ Children in 2nd and 8th grade had higher rates of eczema compared to national average.^**
Nakamura et al (2016)	Great East Japan Earthquake and Tsunami (2011)	Assess impact of disaster on incidence of acute decompensated heart failure	Retrospective chart review	1,733	Coastal and inland areas of Iwate Prefecture (16 municipalities)	Up to 33 months postdisaster
Key findings: A higher degree of tsunami flooding (>10%) was associated with an increased standardized incidence rate of acute decompensated heart failure compared to predisaster levels.^** The percent of refugees in the population was also associated with a higher incidence.^* Seismic intensity had no impact on rates of heart failure.
Nishikawa et al (2015)	Great East Japan Earthquake and Tsunami (2011)	Assess whether periodic hospital visits helped control glycemic levels in type 2 diabetics	Monthly visits	58	Disaster-affected patients attending Soma Central Hospital diabetes outpatient department 3 months before the disaster	Up to 6 months postdisaster
Key findings: Patients with periodic hospital visits postdisaster did not exhibit deterioration in their HbA1c values compared to predisaster (−0.2%).^*
Takahashi et al (2016)	Great East Japan Earthquake and Tsunami (2011)	Assess changes in atherosclerotic risk factors associated with relocation (REL) after disaster	Self-reported surveys and health check-ups	3,160 in REL group; 3,368 in non-REL group	Residents 18 or older in 3 cities heavily damaged by disaster	8 and 18 months postdisaster
Key findings: Relocation was associated with increased body weight (0.31 kg) and waist circumference (0.58 cm), as well as decreased high-density lipoprotein cholesterol levels (−0.96 mg/dL) comparing 18 months (mean) postdisaster to 8 months (mean) postdisaster.^*
Tsubokura et al (2013)	Great East Japan Earthquake and Tsunami (2011)	Assess various metabolic factors postdisaster and compare changes in those evacuating due to tsunami vs radiation disaster	Health screenings in 2010 and 2011	108 in tsunami group; 92 in radiation group	Individuals living in temporary housing in Soma postdisaster	6 to 7 months pre- and postdisaster
Key findings: Patients affected by tsunami demonstrated significant increase in HbA1c (0.3%) and triglycerides (0.17 mmol/L) comparing postdisaster levels to predisaster levels.^ Across entire cohort, BMI, waist circumference, HDL cholesterol, and HbA1c deteriorated comparing pre- and postdisaster levels.^
Yamaki et al (2014)	Great East Japan Earthquake and Tsunami (2011)	Assess change in incidence of acute myocardial infarction (AMI) postdisaster	Physician-reported surveys after AMI event	3,068	General population in 6 districts of Fukushima Prefecture visiting 36 participating hospitals	1 year predisaster to 2 years postdisaster
Key findings: In the year of and year after disaster, there was no appreciable change in overall incidence of AMI compared to the previous 2 years. In 1 region (Iwaki), AMI incidence increased in the 4-12 months postdisaster compared to predisaster levels (38.7 per 100,000).^**
Yokomichi et al (2016)	Great East Japan Earthquake and Tsunami (2011)	Assess postdisaster changes in BMI and prevalence of overweight and obese individuals	Questionnaire and record collection	Cohort study, 7,307; ecological study, 623,220	Children in affected prefectures (Fukushima, Miyagi, and Iwate) and unaffected prefectures	Up to 19 months postdisaster
Key findings: More than 6 months postdisaster, there was evidence of increasing child BMI primarily in tsunami-stricken districts, as well as changes in prevalence of overweight and obese children, although results are not consistent between different prefectures, genders, and age groups.
Yoshimura et al (2016)	Great East Japan Earthquake and Tsunami (2011)	Assess factors affecting physical activity in the elderly postdisaster	Health check-up and survey	4,316	Residents 65 years and older in 4 municipalities in Iwate Prefecture	6-11 months postdisaster
Key findings: Displacement from one's residence and lack of social connections was associated with decreased physical activity among elderly survivors (Displacement: OR 1.37 for men, 1.30 for women; Social connections: OR 1.71 for men, OR 1.79 for women).^ Non-working status was also associated with decreased physical activity.^

Note: ^* p < 0.01, ^** p < 0.05.

Abbreviations: GEJE, Greater East Japan Earthquake; BMI, body mass index; HbA1c, Hemoglobin A1C; AMI, acute myocardial infarction; REL, relocation group; NCD, noncommunicable disease; OR, odds ratio; RR, risk ratio; CEA, carotid endarterectomy; HDL, high density lipoprotein; SDQ, Strengths and Difficulties Questionnaire; RBCs, red blood cell count; Hb, hemoglobin; Ht, hematocrit

Quality of Research Methodology

Control Group and Predisaster Data

Sixteen studies made pre- and postdisaster data comparisons.^61-76 Eight studies used control groups.^{68,69,71,76,78,79,81,83} Four studies included both of these elements^68,69,71,76 (Table 2).

Table 2.

Review of Studies by Design, Control Group, Use of Predisaster Data, Sampling, Validity, and Bias

Citation	Study Design, Control Group,^a Use of Predisaster Data^b	Sampling and Response Rate	Evidence of Validity	Quality Considerations (bias and limitations)
Erskine et al (2013)	Cross sectional; control (-); predisaster (+)	Convenience sample; RR not reported	Not reported	Selection bias given recruitment location
Huang et al 2016	Retrospective chart review; control (-); predisaster (+)	Assessed hospitalizations using data from Urban Employee Basic Health Insurance Program during study period. Discussed likely high response rate/incentive for patients to report hospitalization given high reimbursement rates.	Precise definitions of ischemic heart disease using International Statistical Classification of Disease and Related Health Problems, 10th revision (ICD-10). Used China Earthquake Administration's method to determine hard-hit vs extremely hard-hit areas.	Acknowledge likelihood of good hospitalization reporting rate given 80% reimbursement in study area. Address the following limitations: absence of outpatients from study, absence of control group, and inclusion of numerous confounding variables.
Kamoi et al (2006)	Prospective; control (-); predisaster (+)	Convenience sample; RR N/A	Provide detailed process for BP measurement. Informal but well-established clinical markers for nephropathy, ophthalmopathy, and neuropathy. No validated measure for housing status and damage assessment. Utilized previously used DSM-IV PTSD questionnaire.	Bias not formally addressed
Kamoi et al (2006)	Prospective; control (-); predisaster (+)	Convenience sample; RR N/A	Laboratory tests conducted using previously referenced methods. Provide detailed process for self and clinic BP measurement. Used JSH guidelines for high BP threshold. Unreferenced clinical markers for ophthalmopathy and neuropathy, but referenced for nephropathy. No validated measure for housing status and damage assessment. Utilized previously used DSM-IV PTSD questionnaire.	Bias not formally addressed
Khashram et al (2016)	Retrospective chart review; control (N/A); predisaster (+)	All patients undergoing carotid endarterectomy during the study period	Well-established parameters for diagnosing and retrospectively identifying events (Australasian Vascular Audit, surgeon logbooks, International Classification of Disease coding)	Small sample size
Nakagawa et al (2009)	Retrospective, comparative/descriptive; control (+); predisaster (+)	Examined death certificate data for entire population in Niigata Prefecture 5 years before and 3 years after the disaster; RR N/A	Defined death from AMI according to the International Classification of Disease 10th revision, codes I21 and I22. Detailed process and calculations of death rates using population estimated from national census.	Examined secular trends. Did not adjust for differential degree of damage across municipalities.
Omote et al (2013)	Cross sectional; control (-); predisaster (+)	Assessed all individuals getting health screenings in the year before and after disaster. Predisaster screening rate (RR proxy) 47.5%	Explicit validation of indices to measure degree of damage conducted in partnership with University of Noto Peninsula seismological department. Laboratory values measured based on standards set by Project on Assurance and Standardization of Clinical Laboratory Quality. Used WHO standards for Hb.	Limited physical exam data
Sun et al (2013)	Cross sectional; control (-); predisaster (-)	Multi-stage stratified cluster sampling; RR not reported	Report generally on training of physicians to fill out surveys but did not reference validated tools for other measures. Used WHO/ISH guideline for hypertension	Except for diagnosis of hypertension, specific definitions and evaluation tools for various nonquantitative variables are not specified. Casualties, property loss, and changes of living environment were used as proxies for mental stress, and although evaluation instruments were not discussed or evaluated, the authors do report lack of statistical significance between these factors and development of hypertension.
Wen et al (2012)	Cross sectional; control (-); predisaster (-)	Simple random sample to determine counties, towns, and villages, convenience sampling for households; RR 91.3%	Survey designed by multidisciplinary team. Author-determined list of chronic diseases diagnosed by physicians in hospitals or clinics. Interviewers trained to administer questionnaire in pilot and final phase of study. Two-week morbidity measured by author-specified item. No measure specified for differentiating hard-hit vs less-hit areas.	Address selection bias via multistage sampling. Sampling methodology and sample size contributes to improved generalizability.
Keskinen-Rosenqvist et al (2010)	Retrospective and cross sectional; control (+); predisaster (-)	Convenience sample; RR 45%	Questionnaire developed and adjusted by collaboration between authors and National Centre for Disaster Psychiatry at Uppsala University and the Department of Medical Epidemiology and Biostatistics at the Karolinska Institutet. Physical symptoms assessed using Stockholm Somatic Symptom Checklist used previously in Swedish surveys. Performed confirmative factor analysis to assess construct validity. Used General Health Questionnaire to assess psychological distress, previously used in research. Used Impact of Events Scale-Revised to assess posttraumatic stress symptoms.	Evaluated nonresponders. Consistency with previous literature not formally addressed.
Turner et al (2009)	Cross sectional; control (+); predisaster (-)	Random sample to select households, thereafter convenience. 100% RR for contacted households	Pre-piloted health survey. Briefly address validity of self-reported measures of health.	Power analysis (90% for alpha = 0.05). Acknowledge possible selection bias and lack of generalizability. Address likely recall bias. Specific medical conditions not assessed.
Furusawa et al (2011)	Cross sectional; control (-); predisaster (-)	Convenience sample; RR not reported	Provide detailed information on evaluation tools to assess anthropometric and biochemical measurements.	Bias not formally addressed.
Hagiwara et al (2016)	Cross sectional; control (+); predisaster (-)	Convenience sampling; RR 24.5% during phase 1 and 34.8% during phase 2	Used Athens Insomnia Scale and K6 questionnaire for insomnia and psychological outcomes. No use of validated tool for primary outcome.	Several important variables not measured.
Hasegawa et al (2015)	Retrospective chart review; control (-); predisaster (+)	Convenience sample; RR N/A	Well-defined clinical criteria established for measurements of neuro-otologic disease. At least one measure defined by professional organization. Do not specify measures of mental illness.	Briefly discuss self-selection bias. Given location of hospital within disaster-affected area, assert population likely representative of disaster impact on otorhinolaryngology.
Kikuya et al (2015)	Cross sectional; control (-); predisaster (+)	Convenience sample; RR 39.1%	Referenced specific criteria for threshold responses to determine disease status used in previous literature.	Address self-selection bias
Konno et al (2013)	Observational study; control (+); predisaster (+)	Convenience sample; 58.3% RR for public employees, 51.3% RR for general population	Provide detailed information regarding tools used to collect anthropometric data, blood pressure, and biochemical analyses. Used Hamilton Depression Rating Scale and Brief Job Stress Questionnaire from previous research.	Discuss loss to follow up characteristics. Discuss self-selection/nonresponse bias. Acknowledge absence of job-stress data in general population.
Leppold et al (2016)	Retrospective cohort; control (+); predisaster (+)	Retrospectively identified eligible patients by examining billing records and medical charts	Used Glycohemoglobin Standardization Program for measuring HbA1c values and previously used threshold of 7.0% to define poor glycemic control. Measured residence via average land price in area of residence	20.8% lost to follow up; acknowledge impact on power. Those lost to follow up were analyzed and found to be similar to analysis group, suggesting minimized selection bias. Possible misclassification of evacuation status.
Miyashita et al (2015)	Cross sectional; control (-); predisaster (+)	Convenience sample; RR 30.6%	Used standardized and previously validated questionnaires for assessing allergic symptoms, the International Study of Asthma and Allergies, and the Strengths and Difficulties Questionnaire (SDQ) for assessing social difficulties experienced by children	Discuss self-section bias
Nakamura et al (2016)	Retrospective chart review; control (-); predisaster (+)	Assessed cases of heart failure at all hospitals in study area; RR N/A	Visited all hospitals in study area to review charts to ensure inclusion of all relevant events. Used “tsunami-flooded area” measure to define high- vs low-impact areas, used in previous literature. Used Framingham diagnostic criteria to determine heart failure hospitalization diagnosis.	Acknowledge loss of data from several hospitals
Nishikawa et al (2015)	Retrospective case series; control (-); predisaster (+)	Convenience sample; 7.6% of all patients attending clinic predisaster	Disaster certificates issued by government as evidence of impact. Measure of periodic visitation to hospitals assessed by undefined measure. Used Japanese Diabetes Society values converted to HbA1c National Glycohemoglobin Standardization Program and International Federation of Clinical Chemistry values.	Acknowledge limitations in generalizability and size of cohort. Acknowledge but do not formally assess power.
Takahashi et al (2016)	Longitudinal survey in cohort; control (+); predisaster (-)	Convenience sample; RR 24%	Provide detailed information regarding tools used to collect anthropometric data, blood pressure, and biochemical analyses. Used tools from previous research for several measures (Kessler Psychological Distress Scale, Athens Insomnia Scale, physical activity). Discuss validity for relocation measure and discuss limitations in psychological distress tool.	Discuss and evaluate nonresponse bias and loss to follow up characteristics. Discuss inability to control for several variables.
Tsubokura et al (2013)	Retrospective observational study; control (-); predisaster (+)	Convenience sample; RR 32%	Followed national guidelines for sample measurement, used previously established Japanese reference values to set cutoffs for abnormal values and calculate HbA1c.	Missing data. Acknowledge following limitations: small cohort, self-selection.
Yamaki et al (2014)	Prospective cohort; control (-); predisaster (+)	All patients hospitalized for AMI at 36 participating hospitals across 6 districts were enrolled; RR N/A	AMI diagnosis based on WHO multinational monitoring of trends and determinants of cardiovascular disease criteria.	Used hospital registers so may not have precise count of AMI patients, geo-referenced patients by hospitalization and not residence, miscategorization bias addressed.
Yokomichi et al (2016)	Retrospective cohort and ecological; control (+); predisaster (+)	Convenience sample; for prospective study, 8.8% RR among affected children, 12.4% RR among unaffected children	Used Nationwide Nursery School Survey on Child Health. Detailed process for anthropometric measurements. Used WHO standards for overweight thresholds in prospective study vs Japanese Society for Pediatric Endocrinology standards for ecological study.	Lack of generalizability (missing data) clearly defined in limitations.
Yoshimura et al (2016)	Cross sectional; control (-); predisaster (-)	Convenience sample; RR 94.4%	Cross-validated measures of physical activity with various tools and assessed correlation with other measures of physical activity. Assessed social network with Lubben's scale used in previous literature.	No nondisaster comparison area. Need for future studies to assess neighborhood environment more objectively.

Abbreviations: ^a(+) use of control group, (-) no control group; ^b(+) use of predisaster data, (-) no predisaster data; HbA1c, Hemoglobin A1C; RR, response rate; BMI, body mass index; LDL, low density lipoprotein; HDL, high density lipoprotein; ICD-10, International Classification of Disease-10^th revision; WHO, World Health Organization; ISH, International Society of Hypertension; DBP, diastolic blood pressure; SBP, systolic blood pressure; ISAAC, International Study of Asthma and Allergies in Childhood; PTSD, post-traumatic stress disorder; JIS, Japanese Industrial Standards; CK, creatine kinase; CK-MB, creatine kinase-MB; AMI, acute myocardial infarction; RBCs, red blood cell count; Hb, hemoglobin.

Reliability

One study reported a formal measure of reliability (Cronbach's alpha).⁷⁹ No other studies explicitly reported measures of reliability.

Validity

The following forms of face validity were addressed. Eight studies provided details regarding collection methodology and technology used for laboratory and anthropometric measures.^{64,65,68,74,76,77,81,82} Eleven studies used a previously referenced or validated tool for exposure or outcome measurement.^{64,65,67,68,70,72,76,78,79,81,85} Eleven studies used national or international guidelines for exposure or outcome measurement.^{63,65,69,71,72-76,80,82} One study used cross-validation with other methods of measurement of the same outcome or exposure.⁸⁵ Two studies pre-piloted use of a tool for exposure or outcome measurement.^83,84 Two studies used multiple databases or other sources to ensure inclusion of all cases.^66,72 Content validity was addressed via explicit mention of expert convening or consensus to decide on diagnostic criteria or measurement methodology for exposure or outcome in 7 studies.^{62-65,74,79,84} Overall, only 1 study did not address measures of content or face validity.⁶¹ In 8 studies, there was at least 1 outcome or exposure metric for which there was no assessment of either face or content validity.^{61,62,64,65,73,78,80,84} Only 1 study performed a confirmative factor analysis to assess construct validity, and no studies referenced or assessed criterion validity.⁷⁹

Sampling Strategy and Response Rate

Fifteen studies employed convenience sampling, 7 were retrospective data collection efforts using databases or hospital records, and 3 employed random sampling. Four studies had response rates above 50%, and 4 studies had response rates below 25%.^{68,73,76,78,81,83–85}

Bias

Fourteen studies addressed the presence or mitigation of some type of bias, including nonresponse, loss to follow up, missing data, self-selection, reporting, and recall bias.^{61-63,67-70,72,76,79,81–84} Four studies explained how they attempted to minimize biases, including self-selection, nonresponse, and recall bias, via sampling or study structure.^62,63,83,84 Of the 3 studies that addressed nonresponse bias, 1 evaluated characteristics of nonresponders, and of the 3 studies that addressed loss to follow up, 1 evaluated characteristics of those who were lost to follow up.^69,79

Power Analyses

Three studies explicitly mentioned power and/or the impact of loss of data or loss to follow up on this measure.^69,73,81 One study included a formal power analysis.⁸³

Disease-Specific Findings

Articles addressed the following diseases, with some addressing more than 1: cardiovascular disease (n = 9), diabetes (n = 6), risk factors for noncommunicable diseases (eg, BMI, cholesterol levels) (n = 6), other chronic diseases or symptoms (eg, back pain, anemia) (n = 6), pediatric illness (n = 3), allergic symptoms (n = 2), and general chronic disease burden (n = 2). Two studies addressed interventions in a postdisaster setting, with the remainder focusing on evaluating burden of disease and its predictors.^66,73 No studies evaluated cancer or respiratory disease, 2 of the 4 categories designated as primary noncommunicable diseases by WHO.⁴² Two studies addressed development of renal disease as a complication of diabetes postdisaster, and none substantially addressed outcomes in patients with preexisting renal disease.^64,65 Twelve studies examined housing, displacement, or housing damage as primary or secondary predictors of noncommunicable disease outcomes.^{64,65,69,70,74,78,80-85} (Table 1).

Cardiovascular Disease

Studies addressed the long-term impact of earthquakes and tsunamis both on acute cardiovascular events (eg, myocardial infarction) and chronic conditions (eg, heart failure).^63,71,72,75 In multiple studies, risk factors for hypertension were identified, with public employment, age, family history of hypertension, diabetes, BMI, waist-hip ratio, sleep quality, and blood glucose identified as risk factors for the development of hypertension or increases in blood pressure.^68,80 In one study, morning home blood pressure measurements were lower than clinic blood pressure measurements throughout the study, including predisaster.⁶⁵ After the same disaster, acute myocardial infarction increased 4 to 12 months postdisaster in some geographic areas and not others; acute decompensated heart failure increased for 3 years postdisaster in high-impact areas but remained at predisaster levels in low-impact areas.^72,75 Hospitalizations for ischemic heart disease, defined in 1 study as inclusive of both acute and chronic cardiac conditions and/or complications, increased for 5 years postdisaster but decreased to predisaster levels after 6 and 7 years.⁶³ Of the 2 studies examining the impact of noncommunicable disease management postdisaster, 1 observational study determined that a protocol established predisaster for carotid endarterectomy could be regularly and increasingly administered in a postdisaster setting.⁶⁶ One study evaluated prevalence of hypertension in postdisaster communities but provided no predisaster baseline for comparison.⁷⁷

Diabetes

The potential for disasters to lead to poorer glycemic control, both as a result of environmental factors and stress, has been previously explored.³⁶ In this review, 4 studies found increased HbA1c values postdisaster, although 3 were <0.5%.^64,65,69,82 One study found that the proportion of patients with HbA1c >7.0% increased by 9.5% comparing pre- and postdisaster levels.⁶⁹ In the same study, age, female gender, and rural residency were associated with lower HbA1c values.⁶⁹ One study evaluated the prevalence of diabetes in postdisaster communities but provided no predisaster baseline for comparison.⁷⁷ Of the 2 studies examining the impact of management on noncommunicable diseases postdisaster, 1 found that patients who attended periodic visits for 12 months postdisaster did not experience deterioration of HbA1c values, although this study did not explicitly define the periodic visit measure.⁷³ Of the articles examining changes in HbA1c values, the postdisaster timeframe was 6 to 12 months.^64,65,69,82

Allergic Symptoms

Two studies, both after the Great East Japan Earthquake, addressed the impact of the disaster on allergic symptoms and provided different analyses from the same data collection effort.^67,70 In 1 study of schoolchildren in grades 1 to 8, younger age, history of hospitalization, and worse scores on a social difficulties questionnaire were associated with increased allergic symptoms, including current wheeze or asthma and severe wheeze or asthma with few exceptions among grades.⁷⁰ Both studies determined that prevalence of eczema symptoms was higher than the Japanese average, although the results were statistically significant only for those in 2nd and 8th grades.^67,70

Noncommunicable Disease Risk Factors

Several studies addressed changes in risk factors for noncommunicable diseases, including: (1) weight-related measures (eg, BMI, waist circumference); (2) serum lipid levels (eg, total cholesterol, triglycerides); (3) physical activity; and (4) smoking. Comparing pre- and postdisaster levels, results generally indicated increases in BMI among both children and adults.^76,81,82 Among adults specifically, BMI increased by 0.2 kg/m², body weight by 0.31 kg, and waist circumference by 0.58 cm.^81,82 Results were more variable among children, with 1 study reporting increases in BMI among boys and girls (0.137 kg/m² and 0.200 kg/m², respectively) in the Fukushima Prefecture 19 months postdisaster and decreases in BMI among boys and girls (0.218 kg/m² and 0.082 kg/m², respectively) in the Miyagi Prefecture 19 months postdisaster.⁷⁶ Two studies reported significant changes in serum lipid levels, including decreased high-density lipoprotein cholesterol levels (0.96 mg/dL) and increased triglycerides (0.17 mmol/L).^81,82 Other important risk factors were also addressed, including physical activity and smoking. In Christchurch, 24% of ex-smokers had resumed smoking, and 34.1% of current smokers had increased consumption 15 months postdisaster.⁶¹ After the Great East Japan Earthquake, elderly patients who had been displaced, lacked social connections, and/or were not working had lower rates of physical activity.⁸⁵

Other Diseases and General Burden

Several other chronic diseases outside the WHO noncommunicable disease designations were addressed by studies included in this review, including increases in neuro-otologic diseases and low back pain.^62,78 More generally, chronic disease prevalence and morbidity comparing harder-hit areas to less-hit areas was higher 3 years after disaster, and 2 studies demonstrated increases in chronic nonspecific symptoms more than 1 year postdisaster.^79,83,84

Discussion

The articles examined in this review represent a broad array of noncommunicable diseases that progress and often develop over several months to years after an earthquake or tsunami and focus primarily on evaluating the burden of disease.

While cardiovascular disease was the most well represented noncommunicable disease, results were variable in demonstrating a clinically significant, protracted impact of disaster. Existing evidence strongly supports the hypothesis that acute physiological or psychological stress after disaster leads to increases in negative cardiovascular events such as myocardial infarction and arrhythmias.^35,37 The results of our review suggest that the protracted effects on some cardiovascular disease outcomes—myocardial infarction and acute heart failure exacerbations—may persist for several years after a disaster.^71,72 Studies of hypertension rarely compared pre- and postdisaster data, and in the instances where comparisons were made, data were not collected beyond 8 months postdisaster.^65,68 Data from 1 study on carotid endarterectomy protocol execution suggests that interventions can be reliably administered postdisaster.⁶⁴

Results also supported increases in nonspecific chronic physical symptoms following a disaster. This finding is unique compared to previous studies, which have focused primarily on the WHO noncommunicable disease designations.^24,38,39 Given recent data on the global impact of low back pain as the main cause of disability worldwide, the impact of chronic nonspecific symptoms cannot be ignored.⁸⁶

According to the studies in our review, more psychologically mediated noncommunicable disease risk factors like smoking and physical activity worsened following an earthquake or tsunami.^61,85 These results suggest that postdisaster deteriorations in mental health outcomes may have an overflow effect on noncommunicable disease outcomes.

No articles addressed respiratory disease or cancer, or substantially addressed renal disease, which have unique and intensive resource and personnel requirements that necessitate specific consideration in preparedness efforts (eg, respiratory failure and the need for supplemental oxygen, medications for long-term chemotherapy or radiation treatment, and resources and personnel to administer dialysis). Few articles addressed the benefit of various treatment modalities postdisaster.^66,73 Because disasters can result in variable levels of resource limitation depending on several factors (eg, preexisting stock, severity of disaster, level of preparation, geographic distribution of various healthcare and administrative personnel), it is important to understand whether existing patterns of care are feasible and effective in postdisaster settings or if new standards can be an effective bridge to reestablishing predisaster resource levels. Although issues with the medication supply chain and the impact of disaster on critical medical infrastructure have been identified as relevant to noncommunicable disease outcomes, no literature was identified that linked these issues to noncommunicable disease outcomes in the earthquake or tsunami setting.^19,22,87

This review builds on existing literature by examining noncommunicable disease outcomes in the form of exacerbation, deterioration, or development of new disease. Literature examining the impact of other types of disasters, such as Hurricane Katrina, has supported the prevalence and severity of chronic disease needs in the postdisaster setting and led to the proposal of a network-based model to ensure continuity of care after natural disasters.^{18,19,21,30,53,55,88,89} Overall, our results demonstrate the following when considering noncommunicable disease diagnosis and management post-earthquake or -tsunami: (1) existing conditions may worsen and potentially improve or normalize over time; (2) new conditions may increase in prevalence or incidence or may follow the same pattern; (3) changes demonstrated in the literature may not be clinically significant; and (4) there are often specific variables that place populations at higher risk of experiencing negative noncommunicable disease–related outcomes. Our study is unique in its focus on chronic disease outcomes extending into the recovery period.

While some studies suggest mechanisms underlying the types of longer-term postdisaster changes in physiologic parameters and disease outcomes explored in this review, they are not completely understood.^35,37 It has been suggested that psychological stress from the event itself, personal loss, and displacement contribute to an array of worsening health outcomes.⁷⁷ Indeed, of the articles we reviewed, almost half measured displacement, relocation, or property damage as a primary or secondary predictor of noncommunicable disease outcomes, suggesting that this is an important factor to consider to understand how resulting psychosocial outcomes may modulate, or spill over, to noncommunicable diseases.^{64,65,69,70,74,78,80-85}

Recommendations

Our findings demonstrate that postdisaster noncommunicable disease considerations extend beyond acute exacerbations and include the development of new disease and complex, disease-specific temporal patterns of clinical and laboratory-based deterioration and improvement. In light of these findings, we make the following recommendations for practitioners and researchers:

• Plan for noncommunicable disease exacerbations and shifting disease burden postdisaster: Noncommunicable diseases should be considered well into the recovery phase postdisaster. We recommend that planning and responding agencies account for the longer-term impacts of disaster on noncommunicable diseases demonstrated here. One way to achieve this may be through pre- and postdisaster chronic disease surveillance, which would allow better understanding of local disease burdens in order to inform the most appropriate and efficient responses.¹⁵

• Increase research on postdisaster noncommunicable disease burden, treatment modalities, outcomes, and the effectiveness of existing models for managing noncommunicable diseases after disasters: This review demonstrates that disasters may temporarily alter disease burden compared to projected trajectories. However, data are not robust or generalizable enough to know how long such an increased burden will last or what its magnitude will be. Thus, we recommend that more research be conducted to determine postdisaster noncommunicable disease burden, treatment modalities, and outcomes to aid in planning for changes in noncommunicable disease burden and resource availability postdisaster. We call for this research especially around noncommunicable diseases without any studies identified in this review: cancer, respiratory disease, and kidney disease, a gap that other authors have also recently identified.⁹⁰ This is likely to necessitate establishing long-term postdisaster surveillance of noncommunicable diseases, much like the Fukushima Health Management Survey launched after the Fukushima Dai-ichi Nuclear Power Plant disaster or the formation of the Disaster Cardiovascular Prevention Network.^54,91 Although we call for increased research in this area, we also recognize that several authors have proposed models to address chronic diseases after natural disasters, including a network-based model for institutional coordination and another examining the role for improved public health infrastructure resilience in reducing chronic disease exacerbations after disaster.^21,92 We call for increased implementation and evaluation of these methodologies as well.

• Coordinate response planning regarding noncommunicable disease management with existing healthcare systems: Postdisaster efforts to address noncommunicable diseases will hopefully strengthen healthcare system response to these diseases overall, but predisaster preparation should also include coordination with existing healthcare infrastructure, especially primary health care.⁹³ We recommend coordinated response planning to accommodate changes in postdisaster noncommunicable disease burden and management with existing healthcare systems, especially in the area of primary care, so that redundant efforts are not undertaken and agencies can continue to employ and distribute their resources most efficiently.

Limitations

Our study has several limitations. First, we included only articles originally published in English. As a result, relevant literature may not have been included in this review. Because many noncommunicable diseases take time to develop or worsen over years, literature examining the impact of events from earlier decades may highlight and demonstrate different findings, even though we excluded this literature. In addition, the majority of articles address disasters that occurred in developed countries, with half of the articles addressing 1 disaster specifically: the Great East Japan Earthquake. This may affect the generalizability of our findings, since the disease burden, baseline, and postdisaster resource availability, as well as level and nature of disaster preparedness and response, are likely to be very different between developed countries like Japan and developing countries. Furthermore, due to the varied nature of the included studies, we were unable to categorize studies as occurring in primarily urban or rural areas, even though important differences may exist between these settings. Similarly, given the nature of the studies, we were unable to assess any possible impact of events on older populations, which may face different noncommunicable disease challenges after disasters. Finally, we recognize that some regions remain concerned with earthquake risk only and others are at greater risk for tsunamis. Thus, we separate our presentation of the literature into disaster types (ie, earthquake only, tsunami only, combined earthquake and tsunami). However, we maintain an aggregated approach to analyzing the overall findings, given the small number of studies for each disaster type and absence of clear trends within the literature confined to a single disaster type. Despite the noted limitations, we submit that the overall trends we identify in this study support research on other types of disasters and provide valuable insights in adopting an all-hazards approach to addressing this issue.

Conclusion

This review has demonstrated that earthquakes and tsunamis can have prolonged effects on noncommunicable disease outcomes. Our findings contribute to existing calls for increased research on noncommunicable diseases postdisaster and demonstrate the continued need for additional research to determine the burden of noncommunicable diseases and the effectiveness of various treatment modalities postdisaster, especially in developing countries, and on several important noncommunicable diseases, including cancer, respiratory disease, and renal disease. Varied patterns of postdisaster disease progression make deep, local understanding of predisaster disease burden—via surveillance and coordination with existing primary care providers and systems—essential in conducting preparedness activities to respond to noncommunicable diseases after a disaster.

Footnotes

*

The following search string was used in PubMed and adjusted to the search query syntax required by the other databases: (tsunamis[mesh] OR earthquakes[mesh] OR tsunami*[tiab] OR earthquake*[tiab]) AND (chronic disease[mesh] OR multiple chronic conditions[mesh] OR chronic disease*[tiab] OR multiple chronic condition*[tiab] OR multiple chronic illness*[tiab] OR chronic illness*[tiab] OR post disaster*[tiab] OR non communicable[tiab] OR cardiovascular disease[mesh] OR neoplasms[mesh] OR respiratory tract diseases[mesh] OR diabetes[mesh] OR cardiovascular disease*[tiab] OR neoplasm*[tiab] OR respiratory tract disease*[tiab] OR diabetes[tiab] OR chronic kidney disease-mineral and bone disorder [mesh] OR chronic kidney disease-mineral and bone disorder [tiab] OR renal insufficiency [mesh] OR renal insufficiency [tiab] OR cerebrovascular disorders [mesh] OR cerebrovascular disorder*[tiab] OR wound*[tiab] OR chronic pain [mesh] OR chronic pain [tiab] OR nutrition disorders [mesh] OR nutrition disorder*[tiab] OR renal dialysis[mesh] OR renal dialysis [tiab]).

References

Doocy

, Daniels

, Packer

, Dick

, Kirsch

. The human impact of earthquakes: a historical review of events 1980-2009 and systematic literature review. PLoS Curr, 2013; 5. doi:10.1371/currents.dis.67bd14fe457f1db0b5433a8ee20fb833.

Bartels

, VanRooyen

. Medical complications associated with earthquakes. Lancet, 2012; 379(9817):748-757.

Chan

, Gao

, Griffiths

. Literature review of health impact post-earthquakes in China 1906-2007. J Public Health (Oxf), 2010; 32(1):52-61.

Ahmad

, Ahmad

, Sadia

, Ahmad

. Using selected global health indicators to assess public health status of population displaced by natural and man-made disasters. Int J Disaster Risk Reduct, 2017; 22:228-237.

Gosselin

RA.

Injuries after natural disasters. In: de Dios

, ed. Orthopaedic Trauma in the Austere Environment. Cham: Springer International Publishing; 2016:87-93.

Watson

, Gayer

, Connolly

. Epidemics after natural disasters. Emerg Infect Dis, 2007; 13(1):1-5.

Kouadio

, Aljunid

, Kamigaki

, Hammad

, Oshitani

. Infectious diseases following natural disasters: prevention and control measures. Expert Rev Anti Infect Ther, 2012; 10(1):95-104.

Hasegawa

, Tanigawa

, Ohtsuru

, et al. Health effects of radiation and other health problems in the aftermath of nuclear accidents, with an emphasis on Fukushima. Lancet, 2015; 386(9992):479-488.

Siskind

, Sawyer

, Lee

, et al. The mental health of older persons after human-induced disasters: a systematic review and meta-analysis of epidemiological data. Am J Geriatr Psychiatry, 2016; 24(5):379-388.

10.

Goldmann

, Galea

. Mental health consequences of disasters. Annu Rev Public Health, 2014; 35:169-183.

11.

Nolen-Hoeksema

, Morrow

. A prospective study of depression and posttraumatic stress symptoms after a natural disaster: the 1989 Loma Prieta Earthquake. J Pers Soc Psychol, 1991; 61(1):115-121.

12.

Fergusson

, Horwood

, Boden

, Mulder

. Impact of a major disaster on the mental health of a well-studied cohort. JAMA Psychiatry, 2014; 71(9):1025-1031.

13.

Ardalan

, Mazaheri

, VanRooyen

, et al. Post-disaster quality of life among older survivors five years after the Bam earthquake: implications for recovery policy. Ageing Soc, 2011; 31(2):179-196.

14.

Demaio

, Jamieson

, Horn

, de Courten

, Tellier

Non-communicable diseases in emergencies: a call to action. PLoS Curr, 2013; 5. doi:10.1371/currents.dis.53e08b951d59ff913ab8b9bb51c4d0de.

15.

Slama

, Kim

, Roglic

, et al. Care of non-communicable diseases in emergencies. Lancet, 2017; 389(10066):326-330.

16.

Aebischer Perone

, Beran

. Modifying the Interagency Emergency Health Kit to include treatment for non-communicable diseases in natural disasters and complex emergencies: the missing clinical, operational and humanitarian perspectives. BMJ Glob Health, 2017; 2(1):e000287.

17.

Tonelli

, Wiebe

, Nadler

, Darzi

, Rasheed

. Modifying the Interagency Emergency Health Kit to include treatment for non-communicable diseases in natural disasters and complex emergencies. BMJ Glob Health, 2016; 1(3):e000128.

18.

Hurricane Katrina Community Advisory Group, Kessler

. Hurricane Katrina's impact on the care of survivors with chronic medical conditions. J Gen Intern Med, 2007; 22(9):1225-1230.

19.

Jhung

, Shehab

, Rohr-Allegrini

, et al. Chronic disease and disasters: medication demands of Hurricane Katrina evacuees. Am J Prev Med, 2007; 33(3):207-210.

20.

Ford

, Mokdad

, Link

, et al. Chronic disease in health emergencies: in the eye of the hurricane. Prev Chronic Dis, 2006; 3(2):1-7.

21.

Arrieta

, Foreman

, Crook

, Icenogle

. Providing continuity of care for chronic diseases in the aftermath of Katrina: from field experience to policy recommendations. Disaster Med Public Health Prep, 2009; 3(3):174-182.

22.

Tomio

, Sato

, Mizumura

. Interruption of medication among outpatients with chronic conditions after a flood. Prehosp Disaster Med, 2010; 25(1):42-50.

23.

UN Interagency Task Force on NCDs. Noncommunicable Diseases in Emergencies. World Health Organization; 2016. http://apps.who.int/iris/bitstream/10665/204627/1/WHO_NMH_NVI_16.2_eng.pdf. Accessed December 8, 2017.

24.

Tomio

, Sato

. Emergency and disaster preparedness for chronically ill patients: a review of recommendations. Open Access Emerg Med, 2014; 6:69-79.

25.

Jobanputra

, Boulle

, Roberts

, Perel

. Three steps to improve management of noncommunicable diseases in humanitarian crises. PLoS Med, 2016; 13(11):e1002180.

26.

Mokdad

, Mensah

, Posner

, et al. When chronic conditions become acute: prevention and control of chronic diseases and adverse health outcomes during natural disasters. Prev Chronic Dis, 2005; 2(Spec no):1-4.

27.

Ryan

, Franklin

, Burkle

Jr , et al. Reducing disaster exacerbated non-communicable diseases through public health infrastructure resilience: perspectives of Australian disaster service providers. PLoS Curr, 2016; 8. doi:10.1371/currents.dis.d142f36b6f5eeca806d95266b20fed1f.

28.

Tomita

, Hasegawa

, Watanabe

, Sano

, Hitsuda

, Shimizu

. The Tottori-Ken Seibu earthquake and exacerbation of asthma in adults. J Med Invest, 2005; 52(1-2):80-84.

29.

Kario

, Matsuo

, Shimada

, Pickering

. Factors associated with the occurrence and magnitude of earthquake-induced increases in blood pressure. Am J Med, 2001; 111(5):379-384.

30.

Rath

, Donato

, Duggan

, et al. Adverse health outcomes after Hurricane Katrina among children and adolescents with chronic conditions. J Health Care Poor Underserved, 2007; 18(2):405-417.

31.

Watanabe

, Tani

, Tanaka

, et al. Acute changes in home blood pressure after the Great East Japan Earthquake among patients with chronic kidney disease in Fukushima City. Clin Exp Nephrol, 2013; 17(5):718-724.

32.

Tanaka

, Nakayama

, Tani

, et al. The great East Japan earthquake: blood pressure control in patients with chronic kidney disease. Am J Hypertens, 2012; 25(9):951-954.

33.

Ogawa

, Ishiki

, Nako

, et al. Effects of the Great East Japan Earthquake and huge tsunami on glycaemic control and blood pressure in patients with diabetes mellitus. BMJ Open, 2012; 2(2):e000830.

34.

Aoki

, Fukumoto

, Yasuda

, et al. The great East Japan earthquake disaster and cardiovascular diseases. Eur Heart J, 2012; 33(22):2796-2803.

35.

Kario

Disaster hypertension—its characteristics, mechanism, and management. Circ J, 2012; 76(3);553-562.

36.

Fujihara

, Saito

, Heianza

, et al. Impact of psychological stress caused by the Great East Japan Earthquake on glycemic control in patients with diabetes. Exp Clin Endocrinol Diabetes, 2012; 120(9):560-563.

37.

Kario

, McEwen

, Pickering

. Disasters and the heart: a review of the effects of earthquake-induced stress on cardiovascular disease. Hypertens Res, 2003; 26(5):355-367.

38.

Ryan

, Franklin

, Burkle

Jr , et al. Identifying and describing the impact of cyclone, storm and flood related disasters on treatment management, care and exacerbations of non-communicable diseases and the implications for public health. PLoS Curr, 2015; 7. doi:10.1371/currents.dis.62e9286d152de04799644dcca47d9288.

39.

Miller

, Arquilla

. Chronic diseases and natural hazards: impact of disasters on diabetic, renal, and cardiac patients. Prehosp Disaster Med, 2008; 23(2):185-194.

40.

Kohn

, Eaton

, Feroz

, Bainbridge

, Hoolachan

, Barnett

. Personal disaster preparedness: an integrative review of the literature. Disaster Med Public Health Prep, 2012; 6(3);217-231.

41.

Chaffee

Willingness of health care personnel to work in a disaster: an integrative review of the literature. Disaster Med Public Health Prep, 2009; 3(1):42-56.

42.

World Health Organization. Global Action Plan for the Prevention and Control of NCDs 2013-2020. World Health Organization; 2013. http://www.who.int/nmh/events/ncd_action_plan/en/. Accessed December 12, 2017.

43.

Webster

, Nagler

, Morton

, Masson

. Chronic kidney disease. Lancet, 2017; 389(10075):1238-1252.

44.

Chassin

, Presson

, Seo

, et al. Multiple trajectories of cigarette smoking and the intergenerational transmission of smoking: a multigenerational, longitudinal study of a midwestern community sample. Health Psychol, 2008; 27(6):819-828.

45.

Wickrama

, Conger

, Wallace

, Elder

Jr . The intergenerational transmission of health-risk behaviors: adolescent lifestyles and gender moderating effects. J Health Soc Behav, 1999; 40(3):258-272.

46.

Rimal

RN.

Intergenerational transmission of health: the role of intrapersonal, interpersonal, and communicative factors. Health Educ Behav, 2003; 30(1):10-28.

47.

Carter

, Abnet

, Feskanich

, et al. Smoking and mortality—beyond established causes. N Engl J Med, 2015; 372(7):631-640.

48.

Renehan

, Tyson

, Egger

, Heller

, Zwahlen

. Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Lancet, 2008; 371(9612):569-578.

49.

, Hajifathalian

, Ezzati

, et al. Metabolic mediators of the effects of body-mass index, overweight, and obesity on coronary heart disease and stroke: a pooled analysis of 97 prospective cohorts with 1.8 million participants. Lancet, 2014; 383(9921):970-983.

50.

Aldrich

, Benson

. Disaster preparedness and the chronic disease needs of vulnerable older adults. Prev Chronic Dis, 2008; 5(1):1-7.

51.

Sever

, Lameire

, Vanholder

. Renal disaster relief: from theory to practice. Nephrol Dial Transplant, 2009; 24(6):1730-1735.

52.

Kobayashi

, Hanagama

, Yamanda

, Yanai

. Home oxygen therapy during natural disasters: lessons from the Great East Japan Earthquake. Eur Respir J, 2012; 39(4):1047-1048.

53.

Cefalu

, Smith

, Blonde

, Fonseca

. The Hurricane Katrina aftermath and its impact on diabetes care: observations from “ground zero”: lessons in disaster preparedness of people with diabetes. Diabetes Care, 2006; 29(1):158-160.

54.

Kario

, Nishizawa

, Hoshide

, et al. Development of a disaster cardiovascular prevention network. Lancet, 2011; 378(9797):1125-1127.

55.

Williams

, Kaufman

, Hayden

, Robinson

, Tai

. Comprehensive cancer control in the eye of Hurricane Katrina. Prev Chronic Dis, 2009; 6(4):A139.

56.

Imamura

, Ueno

. Health services: how can we address cancer care after a natural disaster?. Nat Rev Clin Oncol, 2011; 8(7):387-388.

57.

Kishimoto

, Noda

. Diabetes care: after the Great East Japan Earthquake. J Diabetes Investig, 2013; 4(1):97-102.

58.

Khan

, Amatya

, Gosney

, Rathore

, Burkle

Jr . Medical rehabilitation in natural disasters: a review. Arch Phys Med Rehabil, 2015; 96(9):1709-1727.

59.

Harris

, Helfand

, Woolf

, et al. Current methods of the US Preventative Services Task Force: a review of the process. Am J Prev Med, 2001; 10(3 Suppl):21-35.

60.

Akers

, Aguilar-Ibanez

, Sari

AB-A

, et al. Undertaking Systematic Reviews of Research on Effectiveness. Centres for Review and Dissemination Report No. 4. 2d ed. York, England: NHS Center for Reviews and Dissemination, University of York; 2001.

61.

Erskine

, Daley

, Stevenson

, Rhodes

, Beckert

. Smoking prevalence increases following Canterbury earthquakes. Sci World J, 2013; 2013:596957.

62.

Hasegawa

, Hidaka

, Kuriyama

, et al. Change in and long-term investigation of neuro-otologic disorders in disaster-stricken Fukushima prefecture: retrospective cohort study before and after the Great East Japan Earthquake. PLoS One, 2015; 10(4):e0122631.

63.

Huang

, Huang

, He

, et al. Changes in hospitalization for ischemic heart disease after the 2008 Sichuan Earthquake: 10 years of data in a population of 300,000. Disaster Med Public Health Prep, 2016; 10(2):203-210.

64.

Kamoi

, Tanaka

, Ikarashi

, Miyakoshi

. Effect of the 2004 Mid Niigata Prefecture earthquake on glycemic control in type 1 diabetic patients. Diabetes Res Clin Pract, 2006; 74(2):141-147.

65.

Kamoi

, Tanaka

, Ikarashi

, Miyakoshi

. Effect of the 2004 Mid-Niigata Prefecture Earthquake on home blood pressure measurement in the morning in type 2 diabetic patients. Clin Exp Hypertens, 2006; 28(8):719-729.

66.

Khashram

, Falconer

, Mahmud

, et al. Rapid access carotid endarterectomy: winning the RACE following a natural disaster. N Z Med J, 2016; 129(1443):53-60.

67.

Kikuya

, Miyashita

, Yamanaka

, et al. Protocol and research perspectives of the ToMMo Child Health Study after the 2011 Great East Japan Earthquake. Tohoku J Exp Med, 2015; 236(2):123-130.

68.

Konno

, Hozawa

, Munakata

. Blood pressure among public employees after the Great East Japan Earthquake: the Watari Study. Am J Hypertens, 2013; 26(9):1059-1063.

69.

Leppold

, Tsubokura

, Ozaki

, et al. Sociodemographic patterning of long-term diabetes mellitus control following Japan's 3.11 triple disaster: a retrospective cohort study. BMJ Open, 2016; 6(7):e011455.

70.

Miyashita

, Kikuya

, Yamanaka

, et al. Eczema and asthma symptoms among schoolchildren in coastal and inland areas after the 2011 Great East Japan Earthquake: the ToMMo Child Health Study. Tohoku J Exp Med, 2015; 237(4):297-305.

71.

Nakagawa

, Nakamura

, Oyama

, et al. Long-term effects of the Niigata-Chuetsu earthquake in Japan on acute myocardial infarction mortality: an analysis of death certificate data. Heart, 2009; 95(24):2009-2013.

72.

Nakamura

, Tanaka

, Komi

, et al. Sustained increase in the incidence of acute decompensated heart failure after the 2011 Japan earthquake and tsunami. Am J Cardiol, 2016; 118(9):1374-1379.

73.

Nishikawa

, Fukuda

, Tsubokura

, Kato

, Nomura

, Saito

. Managing type 2 diabetes mellitus through periodical hospital visits in the aftermath of the Great East Japan Earthquake disaster: a retrospective case series. PLoS One, 2015; 10(5):e0125632.

74.

Omote

, Kato

, Kido

, et al. Relationship between the degree of property damage and changes in red blood cells, hematocrit, and hemoglobin among victims of the Noto Peninsula Earthquake. Environ Health Prev Med, 2013; 18(2):151-164.

75.

Yamaki

, Nakazato

, Kijima

, Maruyama

, Takeishi

. Impact of the Great East Japan Earthquake on acute myocardial infarction in Fukushima prefecture. Disaster Med Public Health Prep, 2014; 8(3):212-219.

76.

Yokomichi

, Zheng

, Matsubara

, et al. Impact of the Great East Japan Earthquake on the body mass index of preschool children: a nationwide nursery school survey. BMJ Open, 2016; 6(4):e010978.

77.

Furusawa

, Furusawa

, Eddie

, Tuni

, Pitakaka

, Aswani

. Communicable and non-communicable diseases in the Solomon Islands villages during recovery from a massive earthquake in April 2007. N Z Med J, 2011; 124(1333):17-28.

78.

Hagiwara

, Yabe

, Sugawara

, et al. Influence of living environments and working status on low back pain for survivors of the Great East Japan Earthquake. J Orthop Sci, 2016; 21(2):138-142.

79.

Keskinen-Rosenqvist

, Michélsen

, Schulman

, Wahlström

. Physical symptoms 14 months after a natural disaster in individuals with or without injury are associated with different types of exposure. J Psychosom Res, 2011; 71(3):180-187.

80.

Sun

, Zhou

, Chen

, et al. Clinical characteristics of hypertension among victims in temporary shield district after Wenchuan earthquake in China. Eur Rev Med Pharmacol Sci, 2013; 17(7):912-916.

81.

Takahashi

, Nakamura

, Yonekura

, et al. Association between relocation and changes in cardiometabolic risk factors: a longitudinal study in tsunami survivors of the 2011 Great East Japan Earthquake. BMJ Open, 2016; 6(5):e011291.

82.

Tsubokura

, Takita

, Matsumura

, et al. Changes in metabolic profiles after the Great East Japan Earthquake: a retrospective observational study. BMC Public Health, 2013; 13(1):267.

83.

Turner

, Pathirana

, Daley

, Gill

. Sri Lankan tsunami refugees: a cross sectional study of the relationships between housing conditions and self-reported health. BMC Int Health Hum Rights, 2009; 9(1):16.

84.

Wen

, Shi

, Li

, Yuan

, Wang

. Quality of life, physical diseases, and psychological impairment among survivors 3 years after Wenchuan earthquake: a population based survey. PLoS One, 2012; 7(8):e43081.

85.

Yoshimura

, Ishikawa-Takata

, Murakami

, et al. Relationships between social factors and physical activity among elderly survivors of the Great East Japan earthquake: a cross-sectional study. BMC Geriatr, 2016; 16(1):30.

86.

Hoy

, March

, Brooks

, et al. The global burden of low back pain: estimates from the Global Burden of Disease 2010 study. Ann Rheum Dis, 2014; 73(6):968-974.

87.

Boin

, McConnell

. Preparing for critical infrastructure breakdowns: the limits of crisis management and the need for resilience. Journal of Contingencies and Crisis Management, 2007; 15(1):50-59.

88.

Peters

, Moscona

, Katz

, et al. Natural disasters and myocardial infarction: the six years after Hurricane Katrina. Mayo Clin Proc, 2014; 89(4):472-477.

89.

Sharma

, Weiss

, Young

, et al. Chronic disease and related conditions at emergency treatment facilities in the New Orleans area after Hurricane Katrina. Disaster Med Public Health Prep, 2008; 2(1):27-32.

90.

Ruby

, Knight

, Perel

, Blanchet

, Roberts

. The effectiveness of internventions for non-communicable diseaases in humanitarian crises: a systematic review. PLoS One, 2015; 10(9):e0138303.

91.

Yasumura

, Hosoya

, Yamashita

, et al. Study protocol for the Fukushima Health Management Survey. J Epidemiol, 2012; 22(5):375-383.

92.

Ryan

, Franklin

, Burkle

93.

Redwood-Campbell

, Abrahams

. Primary health care and disasters—the current state of the literature: what we know, gaps and next steps. Prehosp Disaster Med, 2011; 26(3):184-191.