Editorial: Armed conflict, war and public health

Health services planning benefits from epidemiological data, which is too often lacking in the developing world. In addition to the importance of considering the role of gender, cultural values and urbanization, the impact of waxing and waning armed conflict and war have particular effects on health and wellbeing that are especially difficult to study and interpret, but which must be considered. In this issue of the journal, Sabbah et al. (2007) report findings from a cross-sectional survey of self-reported morbidity, lifestyle and socioeconomic status among 524 individuals, aged 14 years and above, residing in south Lebanon. The study, carried out in 2000, found significant differences for various morbidity indicators with regard to gender and location in designated rural and urban areas. Women reported more non-specific somatic and emotional symptoms, including headache, lumbar pain, varicose veins and emotional distress (anxiety and depression). Men, however, were more likely to report hearing disorders and cardio-vascular symptoms. Obesity, varicose veins, visual disorders and anxiety/depression were more common among urban, compared with rural, residents. Unsurprisingly, individuals with health insurance sought health care more often than those without, highlighting the value of examining self-reported health status, rather than clinic-based studies, to gauge population health (Miilunpalo et al. 1997). The authors concluded that (i) the so-called epidemiological health transition (shift in causes of death from acute infectious and deficiency diseases to chronic non-communicable diseases) was by then already well under way in Lebanon, and (ii) their study provided a point of reference for assessing the health impact of the 2006 Lebanon War. The first conclusion is supported by complementary mortality data from other sources and rapid urbanization. For example, a cohort study carried out in 1983–1993 among middle-aged and older populations in an urban area of Lebanon found non-communicable diseases as the leading causes of death, particularly circulatory diseases and cancer (Sibai et al. 2001a,b). Examination of the population-based cancer registry established in 1998 found high rates of different cancer types that more closely resembled patterns in high-income countries than neighbouring developing countries (Shamseddine et al. 2004). The process of urbanization, characterised by an increase of Lebanon’s population living in urban areas in 1970 from 59.5% to 83.1% in 1990 and 86.0% in 2000 (United Nations 2006), is recognized to be an important driver of the epidemiological health transition (Utzinger & Keiser 2006). Their second conclusion refers to experience that was not anticipated at the time of the study, and it is appropriate to critically examine how the findings obtained from this cross-sectional survey may be used to assess and monitor the health impact of the 2006 Lebanon War. The perceived morbidity data reported by Sabbah et al. (2007) are valid and representative for southern Lebanon in 2000. This position is justified on the following grounds. First, the study participants were drawn from a random sample of households using a multi-level cluster sampling technique. Second, each individual was interviewed with a standardized questionnaire. Third, risk factors for four common groups of symptoms were analysed with adequate statistical methods, stepwise logistic regression, to adjust for confounding. Several factors nevertheless limit the utility of these data for explaining the impact on health of the 2006 Lebanon War. First, no mortality data were obtained. These are a particularly important indicator of the impact of armed conflict, and they may be obtained by verbal autopsy, which has been a particularly useful tool in countries with inadequate vital registration systems (Soleman et al. 2006). Second, some chronic diseases are less likely than others, and emotional symptoms are more likely, to show the health impact of conflict and war. Although anxiety and depression were examined in the current study, most of the health problems they investigated were associated with chronic diseases arising from the above-mentioned epidemiological health transition. Third, designating a baseline for studying the impact of armed conflict is problematic, inasmuch as Lebanon was already coming out of a 16-year conflict (1975–1991; Sibai et al. 2001a,b; Shamseddine et al. 2004), and the 2006 war was relatively short. Such comparisons should indicate the significance of the chosen time points and the rationale for selecting particular health problems for comparison, indicating what these comparisons are likely to show. Finally, as the authors acknowledge for alcohol and substance abuse, such health survey data may be methodologically limited by stigma and cultural values, which results in underreporting of certain problems. For example, a study of suicide in Oman (Zaidan et al. 2002) and a review of suicide in 17 predominantly Islamic countries (Pritchard & Amanullah 2007) indicate the value of alternatives to self-report for stigmatized conditions. These points suggest a framework for monitoring the health impact of armed conflict and war. For this study in Lebanon, the assertion of a pre-conflict baseline and comparison points on a time line must acknowledge that they are relative, referring to various stages of an ongoing conflict that may intensify and remit. Recent studies in different settings during conflict or in the post-conflict era have shown that careful planning and vigilant conduct of cross-sectional surveys is indeed feasible. Hence, military personnel and the non-combatant civilian population can be interviewed to assess and monitor the impact of war on health, equity and wellbeing (Larsen et al. 2004; Sodemann et al. 2004; Betsi et al. 2006). Nationwide cross-sectional surveys carried out in Iraq in September 2004 and between May and July 2006 best illustrate this claim (Roberts et al. 2004; Burnham et al. 2006). The devastating effects of armed conflict and war on human health and wellbeing are widely acknowledged, and a comprehensive account of the direct and indirect effects on health is presented in a book entitled War and public health (Levy & Sidel 2000). The most obvious and quantifiable direct consequences are the number of casualties and injuries among military personnel and civilians. In 2002, for example, the World Health Organization (WHO) estimated the number of deaths due to war at 172 000 (WHO 2004). The 2006 Lebanon War, also known as July War (in Lebanon) or the Second Lebanon War (in Israel), was a military conflict between Lebanon and northern Israel, commencing on 12 July 2006 and ending on 14 August 2006, when a ceasefire negotiated by the United Nations became effective. The conflict killed more than 1000 individuals, mostly Lebanese civilians. In addition, almost 1 000 000 Lebanese and 300 000 to 500 000 Israelis were displaced. While most of the Israelis have returned to their homes, parts of southern Lebanon remained uninhabitable; towards the end of 2006 some 200 000 Lebanese were internally displaced or refugees. Although one expects that mortality data are relatively straightforward to obtain, with cross-sectional surveys, the research in Iraq indicates that even that is contentious. In September 2004, 18 months after the invasion of military forces into Iraq, an excess mortality of almost 100 000 had been attributed to the Iraq war (Roberts et al. 2004). By July 2006, the excess mortality had risen to over 650 000 with gun fire remaining the most common cause of death, and car bombing-related death on the rise (Burnham et al. 2006). These estimates, however, are still a matter of debate (Burnham & Roberts 2006; Burnham et al. 2007; Roberts & Burnham 2007). This is not surprising, as the annual death toll from this single conflict exceeds the global estimate that WHO put forward in 2002 (WHO 2004). Armed conflict and war are associated with long-lasting physical and psychological impacts on military personnel and the non-combatant civilian population alike, but these indirect effects are far less often documented, and such data are more difficult to obtain than direct effects on mortality. Emotional problems such as anxiety and depression are particularly noteworthy. These data were collected by Sabbah and colleagues in their cross-sectional survey, providing an opportunity to assess and monitor changes over time. Taking estimated mortality figures and disability weights for various conflict-related morbidity indicators, the WHO estimated the global burden attributable to war in the year 2002 at 6.3 million disability-adjusted life years (DALYs) lost (WHO 2004). This burden is approximately 10% of that of diarrhoeal diseases (62.0 million DALYs) and about half that of the tropical disease cluster consisting of Chagas disease, human African trypanosomiasis, leishmaniasis, lymphatic filariasis, onchocerciasis and schistosomiasis (12.2 million DALYs) (WHO 2004). In view of the ongoing conflict in Iraq and the growing number of conflicts throughout the world (Pedersen 2002), the global burden due to war might now be much higher. There are a host of other indirect health effects from conflict that complicate and exacerbate morbidity and mortality. Experience in Côte d’Ivoire, for example, shows that an armed conflict that broke out in September 2002 resulted in the collapse of the health system in the northern, western and central parts of the country. More than 60% and up to 98% of highly trained health personnel (e.g. medical doctors, nurses and laboratory technicians) fled as a result of the conflict (Betsi et al. 2006). Figure 1 illustrates the situation, by comparing the number of health centres and health personnel prior and approximately 1.5 years after the conflict began. Map of Côte d’Ivoire showing the different urban settings in the central, northern and western part of the country where key informants had been interviewed in April/May 2004, approximately 1.5 years after an armed conflict broke out. The effect of the armed conflict on the health system (number of health centres and health personnel before and after the conflict, including percent reduction) are highlighted. Note that regional designations are descriptive but not constitutional (source: Betsi et al. 2006). Analysis of the epidemiological data obtained by Sabbah et al. (2007) focussed on gender and urbanization, cognizant also of broader implications. Large repositories of population health data in low- and middle-income countries, such as those maintained by the INDEPTH Network (http://www.indepth-network.org/), are also potentially useful for assessing the impact of armed conflict and war on health and wellbeing. The coverage of health demographic surveillance system field sites has grown considerably over the past 10–20 years (INDEPTH Network 2002; Tatem et al. 2006), suggesting their value over time for clarifying the health impact of war. Experiences in Guinea-Bissau concerning the relationship between epidemiology and an armed conflict have clearly shown that counter-intuitive results may be expected. For example, a marked decline in case fatality among hospitalized children below the age of 14 years was observed during an armed conflict from June 1998 to May 1999 (Sodemann et al. 2004). On the contrary, mortality increased in the overall population and among children below the age of 5 years during the first 6 months of the conflict. While the crude mortality rate of the entire population was similar to the pre-war era in the second half of the conflict, continued elevated mortality rates were found for girls, but not for boys (Nielsen et al. 2006). Better understanding from health research of the adverse effects of armed conflict and war may help to mitigate their impact by clarifying the toll on health and life. Insofar as one may hope reasoning of that kind may be persuasive politically, such research may also play a role in preventing war (Levy & Sidel 2000). Population health studies should therefore be encouraged to identify and pursue opportunities to document the effects of war on health, health problems and wellbeing. Efforts are needed to acquire an evidence base from monitoring population health that makes the world healthier – recognizing that in conflicted regions, that also means safer and more peaceful. We thank Dr Thomas Junghanss for inviting us to prepare this editorial for Tropical Medicine and International Health, and Ms Susanne Groener for rapid copy-editing. J. Utzinger is grateful to the Swiss National Science Foundation for sustained financial support (project no. PPOOB–102883).