A geographic analysis of population density thresholds in the influenza pandemic of 1918–19

Background Studying influenza pandemics of the past may offer valuable lessons for preparedness for the next great pandemic [1,2]. In this paper, we analyze the 1918 influenza pandemic, during which up to 100 million people worldwide and 670,000 people in the U.S.A. are estimated to have died [3-6]. India, where over 15 million perished in the short span of one year, was the single worst-affected country in terms of total mortality [7,8]. Factors affecting mortality from the pandemic included a lack of immunity to the virus, which resulted in the infection of higher-than-normal numbers of people, its highly contagious nature [9], and the prior presence of other infections such as tuberculosis or subsequent development of pneumonia as the result of infection [10,11]. Because influenza viruses spread through human contact, geography and population density in particular are potential factors for transmission and, indirectly, human mortality. The aim of this paper is to analyze the role of population density in the influenza pandemic of 1918. Because the statistics on influenza mortality for India are deeply flawed, following Davis [8], we estimate population growth trajectories allowing for a break between 1918 and 1919 to capture population loss from the disease, and use a threshold estimation method to test whether low population density districts in India experienced rates of population loss that were different from high density districts. * Correspondence: 1 Asian Studies Center, Michigan State University, 427 N Shaw Lane, Room 301, East Lansing, MI 48824, USA Full list of author information is available at the end of the article INTERNATIONAL JOURNAL OF HEALTH GEOGRAPHICS © 2013 Chandra et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( reproduction in any medium, provided the original work is properly cited. Chandra et al. International Journal of Health Geographics 2013, 12:9 Epidemiology of the influenza pandemic of 1918–19 in India The 1918–19 influenza pandemic was one of the worst epidemics in history with an estimated global mortality between 20 and 100 million [9,12]. The pandemic occurred in two or three waves [13,14]. The first was a mild wave in the spring of 1918, followed by a second more severe wave in the following autumn that was responsible for the majority of deaths. The third wave was sporadic [15]. The influenza pandemic of 1918–19 in India shared many characteristics with the pandemic in other parts of Asia and the world. In this account, we draw heavily on the Report of the Sanitary Commissioner of India for 1918 [16] and Chandra [17]. The virus is believed to have entered India in the early months of 1918 through the port of Bombay on the west coast. This first wave was relatively mild. The same early and mild first wave was observed in other countries including Indonesia [18], England, Scotland, and Wales [19], Portugal and Spain [20], Mexico [21], and Peru [22], and cities including New York City [23] and Copenhagen [24]. It subsided by August, only to be followed by a second and far more virulent wave that peaked between September and November 1918 in various parts of India. Indonesia [17], England, Scotland and Wales [19], Portugal and Spain [20] and Mexico [21] experienced a similar pattern of timing and relative severity. Populations that were exposed to the virus early have been shown to have benefited from acquired immunity against the deadly second wave in Denmark [24], Norway [25], and Britain [26,27]. A distinctive characteristic of this epidemic was its disproportionate impact on victims aged between 15 and 35 years [6,14,15,28-30]. This phenomenon was also observed in India [16]. Epidemics and urban responses in history It is no coincidence that the public health community worries about new and virulent infectious diseases [31]. As recently as 2009, an episode of pandemic influenza is estimated to have claimed between 151,700 and 575,500 lives worldwide [32]. While vaccinations are frequently the first line of defense against influenza viruses, development of vaccinations for a new strain of virus may take months. In such a situation, only short-term measures including social distancing, and in extreme cases evacuations and quarantines, can protect citizens from a severe epidemic outbreak. The early introduction of social distancing measures, such as school and church closures and banning of mass gatherings, significantly reduced excess mortality during the 1918–19 influenza pandemic [33]. With similar contagious diseases, such as SARS, plague and cholera, quarantines and evacuations have been used to counter severe outbreaks. In the 15th century, for example, the government of Venice combated outbreaks of plague by establishing the “Lazzaretto Vecchio” on a small island off the coast of the Piazza San Marco [34]. More recently, in the late 19th and early 20th centuries, New York City quarantined travelers on Swinburne and Hoffman islands to prevent the spread of cholera [35]. At about the same time, the city established the Metropolitan Board of Health to develop zoning codes to prevent overcrowding in the city and to establish standards for sanitary conditions [36]. With reference to the influenza pandemic of 1918-19 in India, the Sanitary Commissioner of India wrote: “As the striking distance of the influenza virus is probably short the obvious ideal is free ventilation and isolation of sufferers with a view to increase the air space between infected and uninfected” [16, p.66]. Over the past two decades, sudden virus outbreaks that could have led to widespread human pandemics, including H1N1 [37], SARS [38], and H5N1 [39] prompted a series of studies on non-pharmaceutical interventions [40-48], including measures to increase social distance, such as the creation of spatial barriers through quarantine [49,50], relocating populations to ‘safe’ areas [51,52], or imposing travel restrictions [53]. Evacuations, a last resort among social distancing measures, are still used; in the aftermath of the Haiti hurricane of November, 2009, and the subsequent earthquake of January 12, 2010, the government ordered the evacuation of the capital, Port-au-Prince, to prevent the spread of epidemic cholera [54]. In sum, it is widely believed that public health interventions, including social distancing measures and the controlled movement of people to either sequester those infected or as a means to lower population density below some critical threshold can significantly decrease the likelihood of a contagious disease spreading. Yet, to date, there is little if any guidance as to what such a population density threshold might be for any disease. Using the influenza pandemic of 1918 as a case, this paper presents an approach to identifying such a threshold value as a guideline for public health policy. Population density as a factor in influenza population loss Studies examining the potential relationship between population density and mortality during the 1918 influenza pandemic have produced mixed results. Garrett [55] found a positive relationship between mortality rates and population densities measured on a state-wide scale in the USA. Once cities were introduced into the equation, normalizing the mortality rate of the cities with those of the states also showed a positive relationship with population density. For Nigeria, crowding contributed to comparatively higher mortality than less-crowded areas, and “there is enough evidence to support the view that Chandra et al. International Journal of Health Geographics 2013, 12:9 Page 2 of 10 large towns suffered more than small and remote villages” [56]. During the 2009–10 influenza pandemic, it was observed that H1N1 infections were sustained over longer periods of time in Taiwanese areas with higher population densities [57]. Theoretical models for influenza and other transmissible respiratory diseases consisting of agent, host, and environment interactions usually require a high host density [58].