Our worst nightmare may not be a new one.25
Influenza is both familiar and unknown. Although easily distinguished from most common colds by a characteristic moderate to high fever and dry cough, influenza A can exhibit an extremely broad range of symptoms (including sore throat, headache, bone aches, conjunctivitis, dizziness, vomiting, and diarrhea) that overlap with numerous other so-called “grippes, catarrhs and colds.” The continuing, rampant prescription of antibiotics for influenza is proof of the difficulty that most general practitioners and clinic staff face in distinguishing between viral and bacterial infections. “[I]t is now accepted,” writes one world authority, “that influenza is quite protean in its manifestations. Influenza cannot be distinguished readily on clinical grounds from other acute respiratory infections, and during virologically confirmed outbreaks of influenza the proportion of influenzal illnesses confirmed by laboratory tests as being influenza is currently about half.”26
If diagnosis is often mere guesswork, an accurate census of influenza mortality is almost an impossibility: except during pandemics, influenza is usually only the accessory to murder. By destroying the ciliated epithelial cells that sweep dust and germs out of the respiratory tract, flu encourages superinfection by bacteria. (Haemophilus influenzae—widely believed in 1918–19 to be the actual pathogen of the pandemic—is a famous fellow traveler.) A lethal synergy is believed to operate between influenza A and pneumonic bacteria, with Staphylococcus aureus and Strepto coccus pneumoniae being particularly vicious; thus, bacterial pneumonia is the most common, or at least the most clearly associated cause of influenza deaths. But how to distinguish influenza-related cases from the rest of pneumonia mortality? As Registrar General of England William Farr first realized during an influenza epidemic in 1847, the infection’s well-defined seasonality (October to March in the Northern Hemisphere) in temperate countries allows a rough calculation of excess mortality by simple subtraction of the annual average from the winter spike.27
Although epidemiologists now use sophisticated regression modeling, influenza mortality is still estimated in North America and Europe as excess annual mortality. Recently, however, it has become evident that the traditional reporting category “pneumonia and influenza” shortchanges influenza’s deadly impact. Most of the winter spike in ischemic heart disease, diabetes, and cerebrovascular disease mortality may also result from the impact of the annual flu epidemic; conversely, “influenza vaccination has been associated with large reductions in the risks of primary cardiac arrest, recurrent myocardial infection, cardiac disease and stroke.”28 In a normal year, researchers now believe that influenza kills between 36,000 to 50,000 mostly elderly (and especially poor) Americans, a reality that belies the benign image of flu as nothing more than a winter nuisance.29 Sadly, an infection that primarily kills infants and old people is not likely to arouse as much concern as a disease that kills young or middle-aged adults.
As difficult as it is to estimate flu mortality in this country, global influenza mortality is mere conjecture. “There is,” writes one research team, “an under-appreciation and an underestimation of the impact of influenza in the developing world.”30 It is sometimes said that flu kills 1 million people worldwide each year, but the toll could be considerably higher because annual influenza is the least recognized of all so-called “captains of death.” Neither China nor India, for instance, reports flu statistics to the World Health Organization.31 In tropical countries, moreover, the absence of well-defined seasonality in the incidence of influenza makes estimation of excess mortality difficult. This dearth of data, in turn, has reinforced the stereotype that there is no significant influenza burden in Asia or Africa.
While high death rates from acute respiratory infections in the tropics are often attributed to tuberculosis, recent research has established that a majority of acute respiratory deaths are caused by viruses, and that tropical countries have influenza mortality rates at least equivalent to those in the mid-latitudes. Indeed, “infection probably has an even greater relative impact on the health of persons from developing countries who are already susceptible to complications because of underlying malnutrition, tropical diseases and HIV.”32 As studies in Southeast Asia have shown, “overall influenza-associated mortality in a region with a warm climate, such as Hong Kong, is comparable with that documented in temperate regions.” Moreover, infant mortality from influenza is probably considerably higher in low-income tropical countries.33
Influenza is most of all a mystery disease in sub-Saharan Africa. The region is the weakest link in the global influenza-surveillance network coordinated by the WHO: in recent years Côte d’Ivoire, Zambia, and Zimbabwe have closed down their national flu surveillance systems after pleading debt and bankruptcy; currently only South Africa and Senegal actively track flu cases and have the laboratory resources to isolate and characterize subtypes. In the rest of Africa, serious flu cases are commonly conflated with malaria or just added to the “acute respiratory infection” (ARI) grab bag. Yet annual influenza in Africa does often produce explosive local outbreaks, such as the 2002 epidemic in Madagascar which overwhelmed the country’s healthcare system, or the massive irruption six months later in the Equateur Province of the Democratic Republic of the Congo which yielded shocking rates of secondary pneumonia.34
Third World influenza is also largely invisible or poorly studied in the historical record. The apocalyptic pandemic of 1918–19—according to the WHO, “the most deadly disease event in the history of humanity”—is the template for the public-health community’s worst fears about the imminent threat of avian influenza.35 After two generations of cultural amnesia, popular interest in the history and legacy of the “Spanish flu” (so called because uncensored newspapers in neutral Spain were the first to report its arrival) has undergone a dramatic revival in recent years. Since 1974, when Richard Collier published an anecdotal history based on interviews with hundreds of survivors, an impressive succession of historians and science journalists—including Alfred Crosby, Gina Kolata, Pete Davies, and, most recently, John M. Barry—has focused on the far-reaching impacts of the pandemic on American life, medical research, and the outcome of World War I. Several writers have also chronicled the recent expeditions to Alaska and Spitzbergen in the Arctic to try to retrieve the 1918 virus from the frozen cadavers of its victims, as well as the dramatic successes of U.S. Army scientists, led by Jeffrey Taubenberger, in reconstructing much of the 1918 virus’s genome.
The threat of a new pandemic, meanwhile, spurs continuing research into many aspects of the 1918 virus’s molecular structure; the enigmatic circumstances of its emergence (reassortment or recombination?), its geographical origin (a Kansas army base, the trenches in France, and southern China are all proposed epicenters),36 and its distinctive mode of attack (which produced singularly high mortality among young adults). Despite renewed scholarly investigation into the 1918 pandemic, however, shockingly little attention has been paid to the disease’s ecology in its major theater of mortality in 1918–19: British India. This oversight is analogous to the history of the First World War having been written with a vivid, sustained focus on the campaigns in the Balkans and Gallipoli while devoting only an occasional aside or footnote to the slaughter on the Western Front.
Pandemic Mortality 1918–19—Revised37
Worldwide | (a) 21.64 million | (b) Asia48.8 to 100 million |
Asia | 15.78 | 26 to 36 |
India | 12.50 | 18.5 |
China | . . . . . . | 4 to 9.5 |
East Indies | .80 | 1.5 |
Europe | 2.16 | 2.3 |
Africa | 1.35 | 2.38 |
W. Hem. | 1.40 | 1.54 |
USA | .55 | .68 |
(a) Jordan (1927) (b) Johnson & Mueller (2002)
The enormity of influenza’s impact on India has never been questioned. For decades the authoritative guide to worldwide pandemic mortality was the 1927 American Medical Association-sponsored study—Epidemic Influenza—by Edwin Oakes Jordan, editor of the prestigious Journal of Infectious Disease, who had spent years poring over death statistics. The huge spike in mortality during the fall of 1918—U.S. life expectancy fell by ten years—allowed him to make estimates of the pandemic toll despite the absence of influenza data per se (see Table 2.1). Jordan believed that global mortality from influenza was in the range of 20 to 22 million (about 1 percent of the human race), with India alone suffering 12.5 million deaths, almost 60 percent of the total. (U.S. flu deaths, by contrast, constituted only 3 percent of the world total.) But at an international conference on the history of the great pandemic, held at University of Cape Town in September 2001, medical demographers Niall Johnson and Juergen Mueller challenged Jordan’s estimates “as almost ludicrously low.” Reviewing modern research, they came to the conclusion that “global mortality from the influenza pandemic appears to have been of the order of 50 million.” Moreover, the two warned that “even this vast figure may be substantially lower than the real toll, perhaps, as much as 100 percent understated.” In other words, it is possible that mortality was actually closer to 100 million or more than 5 percent of the contemporary world population. In their revision, Indian deaths (mainly in the deadly second wave of influenza after September 1918) are reckoned at 18.5 million, although another scholar thinks 20 million is more likely.38
What explains the extraordinary mortality in India? “Famine and pandemic,” observes I. Mills, “formed a set of mutually exacerbating catastrophes.” Indeed, these two factors were exquisitely synchronized during the fall of 1918. As Mills explains in one of the few scholarly articles on the Indian experience, the milder first wave of the pandemic arrived in Bombay in June (via the crew of a troop transport) just as the southwestern monsoon was failing throughout much of western and central India; the resulting drought led to soaring grain prices and famine conditions in Bombay, the Deccan, Gujarat, Berar, and, especially, the Central and United Provinces. (Although not mentioned by Mills, grain exports to England and wartime requisitioning practices undoubtedly contributed to price inflation and food shortages as well.) In September, as the famine was worsening, the second—more deadly—wave of influenza arrived, again via Bombay.39
What followed was the kind of chain reaction (or positive feedback of disasters) that has become so familiar in the history of the modern Third World. “In Bombay Presidency,” writes Mills, “the severe second [influenza] wave came at the time of the harvest of the early crop, and sowing of the late crop. With morbidity estimated to be in excess of 50 percent of the population, and with the concentration of severe attacks in the most productive age range, 20–40 [years], the effect on agricultural production was extreme.” The area of grain production decreased by one-fifth while staple food prices doubled.40 The “absolute lack of any public health organization redoubled infection’s impact upon the famished population.” The Raj heavily taxed the peasantry to support the Indian Army but spent virtually nothing on rural medicine. (“The Surgeon-General conceded that mortality would have been reduced had it been possible to provide immediate medical aid and suitable nourishment to those attacked.”)41 The American missionary Samuel Higginbottom, who was director of agriculture in the state of Gwalior, wrote to a friend that “influenza has been fearful. Hundreds of bodies daily floating in the river. No official figures have been published for India as a whole, but in villages in Gwalior State that are under my charge the death rate during October and November was from 20 to 60 percent. Cholera, plague, and other epidemics from which India suffers have never shown such a death rate as Influenza.”42
Desperate refugees from the countryside flooded into the slum districts of Bombay and other cities; there, influenza cut them down by the tens of thousands, “like rats without succour,” according to the nationalist paper Young India.43 Mortality, Mills emphasizes, was strictly “class oriented,” with almost eight times as many deaths among low-caste people in Bombay as among Europeans or wealthy Indians—the poor seemed to have been the victims of a sinister synergy between malnutrition, which suppressed their immune response to infection, and rampant bacterial pneumonia.44 Outside of the crowded urban slums, flu mortality was generally highest in the famished west of India rather than in the east, where the crops had not failed.
Presumably hunger played a similar role in influenza mortality in China, the East Indies, and even Germany, where the Allied blockade had reduced the caloric intake of the urban poor, especially women and children, to dangerous levels. Certainly, every writer on the pandemic has noted its particular affinity for poverty, substandard housing, and inadequate diets. The slum districts of port cities, from Boston to Bombay, seemed to offer especially favorable conditions for spread of the pandemic in its more virulent form.45
The pandemic also formed lucrative partnerships with other epidemic diseases. Iran was a grim case in point: according to a careful study by historian Amir Afkhami, the nation of 11 million suffered the greatest relative mortality of any major country, between 8 and 22 percent of the total population. The pandemic hitchhiked the military supply route from Bombay to the British occupation force in this supposedly neutral country. Iran was already reeling from several years of drought, famine, cholera outbreaks, and the depredations of marauding armies. In addition, the British had callously aggravated the famine by requisitioning the grain surplus from the large estates, leaving little for a hungry population. Writes Afkhami,
At the dawn of influenza’s outbreak in Iran in the spring of 1918, grain supplies were at a low point, and prices had already more than doubled from the preceding six months (when they had reached a ten-year peak). This scarcity continued even following the spring harvest, and villagers, especially in the southern and central provinces, were scarcely surviving on millet-meal and berries. . . . As if starvation were not enough, in 1918 the Iranian people also had to grapple with a widespread typhus epidemic, which was taking its toll in both urban and rural areas. Consequently, the flu came into an environment already beset by the calamities of war, famine and disease.46
But Akfhami argues that the principal multiplier of influenza mortality in Iran, even more than hunger, was malaria. He finds dramatic correlations between malaria incidence and influenza mortality, both among the local population and the Indian troops of the British Army. Cities with chronic malaria, such as Mashhad, had influenza death rates triple those of cities with low malaria rates, such as Tehran. The climax of pandemic mortality in November coincided with the usual “peak period of malignant tertian malarial fevers among Iranians.” Akfhami also observes that malaria sufferers, including both Iranians and Indians, were afflicted with anemia and were notoriously susceptible to pulmonary infections.47
Poverty, malnutrition, chronic illness, and co-infection were thus powerful determinants of the precise tax that the 1918 influenza exacted from different populations. Indeed, the global pandemic itself was really a constellation of individual epidemics, each shaped by local socioeconomic and public-health conditions. In some countries, such as India and Iran, the co-factors (hunger, malaria, anemia) formed deadly nonlinear synergies with influenza and its secondary infections. Although most of the literature on the 1918 pandemic has focused on its unusual preference for young adults, including the robust and well-fed young soldiers of the American Expeditionary Force in France, the correlation between social class and lethality in virtually every country was no less striking. In the most sophisticated analysis of pandemic mortality yet undertaken—a case-study of the 1918 virus in Sydney—Kevin Cracken and Peter Curson found that “the working class and blue-collar workers experienced the heaviest death rates,” particularly in the inner city, and that unemployment was as consistent a predictor of mortality as more conventional epidemiological factors such as persons per room density.48