Daily Newspaper View of Dengue Fever Epidemic, Athens, Greece, 1927–1931

Controversy remains about possible dengue hemorrhagic fever during the epidemic.

T he dengue fever (DF) epidemic of 1927-1928 in Greece, which fi rst affected the Athens area, was the most recent dengue epidemic in Europe. Transmitted by Aedes aegypti mosquitoes, the epidemic probably involved dengue virus type 1 (DENV-1) and type 2 (DENV-2) (1-4). The rapid economic and social development of the continent has since led to dramatically reduced habitats of the vector and, most likely, to its elimination and is the main reason the likelihood of a similar event is small. Yet, the recent invasion of Europe by Ae. albopictus mosquitoes (5) and the marked ecology of this mosquito make emergence of a new DF epidemic possible, even if the impact of such an event is not expected to be of the same dimension as the epidemic described here.
Greece in 1927 lagged behind other countries in development. For example, most roads in Athens were unpaved; electric service was intermittent; a citywide sewage system was nonexistent; and the potable water supply was rudimentary, often forcing residents to store water in containers. Moreover, the population of the metropolitan Athens area, like the remainder of the country, had increased markedly because of the exchange of populations between Turkey and Greece after the war between the 2 countries and the defeat of Greece in 1922 (6). The Hellenic National Statistical Agency reported the arrival to Greece of ≈1 million refugees, leading to a 23.68% increase of the population during 1920-1928. In the Athens metropolitan area, the increase was ≈68%, and most of the newcomers were destitute and lived in extremely poor housing conditions (7).
In addition to these problems, the country was deeply divided politically into royalists and liberals, a division initiated early in 1915 by the rift between King Konstantin I and Prime Minister E. Venizelos about whether Greece should enter World War I. This division, exacerbated by the defeat of Greece in the Asia Minor campaign of 1921-1922, contributed to the lack of political solutions for the country's major problems (8).
During this period of hardship, the DF epidemic struck Greece. The epidemic was then often called the dengue pandemic because almost the entire population of Athens (population ≈600,000) was affected. On the basis of the current defi nition of pandemic, the term epidemic is used here to describe the events of the summers of 1927 and 1928.
By focusing exclusively on all 561 items published in 1 daily newspaper during the epidemic, this review compares reports from 83 years ago with today's knowledge. I chose the fully digitized archives of the Athens newspaper Η Καθημερινή (I Kathimerini) (9). This newspaper was, and still is, one of the major newspapers in Greece (currently fi rst in daily circulation). During the time of the DF epidemic, the newspaper held a steady antiliberal and proroyalist position at a time when the liberal party ruled Greece. This position is often indicated in commentaries about the government's handling of the epidemic, which will only briefl y be mentioned here. Unless specifi cally referenced, all sources of original statements listed below (translated into English and appearing in quotes) are based on the pieces published by I Kathimerini.

1927
I Kathimerini fi rst mentioned DF on November 1, 1927, not as news but rather as a short joke about fi nancial matters, followed several days later by an advertisement for "FLY FUME" (ΦΛΑΪ ΦΙΟΥΜ), which "...protects you from the awful Dengue Fever, because it eradicates immediately the mosquitoes, which transmit it...." While publishing more jokes during the next few weeks and further advertisements for FLY FUME, I Kathimerini did not mention the DF epidemic until August 1928, when an article reported that the epidemic during the fall of 1927 had resulted in 3,000 cases with 0 fatalities; a few days later, the number of ill persons was reported as 75,000. Moreover, the second article suggested that many of these DF cases were initially mistaken for the yearly recurrence of "3-day fever," which was known to be transmitted by phlebotomine sandfl ies. In August 1928, citing an unnamed physician, I Kathimerini reported that the previous year's epidemic had started in mid-August in a middle-class household in the center of Athens through a young woman in whom DF symptoms developed 2 days after she arrived from Alexandria, Egypt.

1928
On August 2, 1928, when the second epidemic had been ongoing for several days, I Kathimerini began publishing frequent articles mentioning DF. An article in early September mentioned that the fi rst 3 DF cases in 1928 had appeared in mid-July in the center of Athens among construction workers. This article also criticized the medical authorities for not issuing directions and for not taking any palliative and prophylactic measures despite the fact that "whole families stay in bed" or "big factories are forced to stop work." The fi rst real medical facts were published several weeks later in late August.
The headline of a long article appearing on August 15 was "25,000 Athenians are burnt by DF" and continued "... science remains speechless in front of the mysterious disease." It described the symptoms of DF and mentioned a common prescription ("one spoon per hour of Riviera Potion 250 gr, Pyramidon 1 gr, Aspirin 1 gr"). The article concluded that "all measures and drugs suggested by [medical] science are symptomatic and not therapeutic." It also reported that "those [persons] injected with 606 and Bismuth did not get DF" (606 was the synonym for arsphenamine/salvarsan).
The Table indicates the numbers of cases in the DF epidemic according to fi gures published by I Kathimerini (no fi gures were published after August 30, 1928). The Athens city center was hardest hit. An article on August 22 reported 30 deaths on the previous day, the fi rst time that I Kathimerini reported deaths resulting from the epidemic (Figure). On that same day, another article noted that the fatal cases resulted from a "[coexistent] different organ affection" [sic]. Isolation of the DF "microbe" was unsuccessful, leading to the opinion that the transmitting mosquito might be poisonous. The vector was identifi ed in different articles as Stegomyia fasciata mosquitoes, an old name for Ae. aegypti mosquitoes.
On August 24, some of the effects of the epidemic on society were reported for the fi rst time. For example, the Athens-Piraeus Railroad used fewer trains, half the judges of the Athens Lower Court were absent, and the armed forces draft was suspended. Although more fatal cases were reported, hemorrhagic DF-like illness was not described, and deaths were attributed to older age or preexisting diseases. Some articles suggested that residents  (8) *Blank cells indicate no information. †Includes the neighboring city of Piraeus (population 250,000) and the remaining suburban area that had a small population in the 1920s.
were afraid the epidemic might be yellow fever, but medical authorities explicitly denied this concern because "no patient ever showed any symptoms of hemorrhage" (the term σύμπτωμα in Greek covers both symptoms and signs). However, Cardamatis (10) described the course of the epidemic on the island of Aegina, located ≈25 km from the mainland, and stated that hemorrhage in all organ systems was the most common complication and cause of death. Furthermore, on August 25, I Kathimerini stated that "the present epidemic is malignant" and that neurologic disorders or cardiomyopathies often developed in the DF patients. To prevent complications, articles suggested that "intravenous salysalic [sic] acid, well sterilized [sic] should prevent cardiac diseases, especially myocarditis, ensuing from DF." Experts from Greece and abroad were asked for their views, but often the same expert cited in different articles in the same issue expressed contradictory opinions (e.g., possibility of DF transmission by house animals, excluded a few pages later). A common theme was that DF was associated with environmental degradation, and general measures for other diseases, such as cholera, should successfully curb the epidemic. The British power company performing public works and digging in Athens was accused of "providing breeding sites" for mosquitoes. Authorities seemed to be aware that only control of mosquitoes would help alleviate the epidemic and suggested that any "ditch should be covered as soon as possible and petroleum and/ or lime should be used in order to prevent mosquitoes from breeding." City authorities were advised to "wash the streets regularly and swamps should be dried the soonest possible." These proposed measures were accompanied by proposals to "brush one's teeth regularly" and take the drug "emostyrol." Finally, several clergymen's suggestion that "prayers and litanies would help end the epidemic" was made public.
On August 25, I Kathimerini reported the fi rst cases outside the Athens area. Persons who had traveled to their birthplaces to vote in the general elections a few days earlier were held responsible for transporting the disease. On August 28, an article appeared that described what was then generally known about DF without citing the source of the information. Until that date, this article was 1 of only 2 scientifi cally sound articles describing the disease and its treatment. Additional articles contained amateur advice about how to cure DF (e.g., by "drinking milk and lemonade"). Development of an effi cient drug also was reported (an "iodine solution per os" without further details), later followed by a letter from the chemist who developed it, cautioning that success achieved might have been purely coincidental.
More scientifi cally sound articles appeared in September. For example, the Minister of Health, an established pediatrician, explained how the "microbe" remained in Greece throughout the year; he had seen 1 DF case himself, 3 months before the new outbreak. He also explained that, after the acute phase, patients were expected to have lifelong immunity against dengue. Dr. Georges Blanc, head of the Hellenic Pasteur Institute, who was in France at the beginning of the epidemic, returned to Athens and granted an interview in which he tried to correct several statements. Moreover, a report signed by him and the health authorities also described facts about DF and its mode of transmission. I Kathimerini published a summary: the vector was Stegomyia mosquitoes, and the "microbe" had no relation to Spirochaeta (an association between DF and syphilis was considered possible by many physicians) or Plasmodium laverani.
One article cited a different newspaper claiming that the epidemic was not in fact DF but, rather, yellow fever. Two unnamed physicians asked for their opinion were cited as saying that if the disease were yellow fever, mortality rates would have been much higher. Initial reports on the epidemic in mid-August claimed that, compared with 1927, the new DF outbreak was much more malignant, on the basis of the lack of deaths reported during the previous year. I Kathimerini  some deaths might have occurred in 1927 but, because of the previous ill health of the patients, could have been attributed to other causes. The second month of the outbreak saw qualitatively different articles. Small news pieces about the everyday effects of the epidemic appeared frequently: entrance examinations to the universities were postponed, the army draft was deferred, theaters were closed, spouses and children were sent to resorts far from Athens, courts were closed because of the lack of judges and jurors, and the mode of taxation was changed to address lower income. Brief articles reported numbers of DF cases from Athens and elsewhere, including non-Greek cities such as Smyrna (now Ismir), Copenhagen, and Berlin, attributed to travelers from Greece. In mid-September, I Kathimerini reported that the epidemic was clearly weaker: "... the crowd at the horse races was again high after several weeks of low attendance." On the same day, a 12-word item bore the headline, "ONLY TWO DEATHS!" On September 22, I Kathimerini published a report of the Ministries of National Economy and Health about the total number of deaths caused by the epidemic, aimed at calming foreigners wishing to travel to Greece. The numbers were corrected slightly a few days later, but the overall picture did not change (Figure). I Kathimerini did not subsequently publish new numbers of fatal cases.
The end of the epidemic saw a new kind of article about a potential reappearance in the following year. In a letter to I Kathimerini, a physician claimed that several patients bitten by infected mosquitoes late in the fall would become sick only in early spring. Therefore, in addition to mosquito control measures, an absolute isolation (i.e., quarantine) of the fi rst patients of 1929 should prevent a new spread.
On October 29, I Kathimerini published a fi nal report of the Nomiatros, the highest medical authority in the Νομός (i.e., county) of Attica; he claimed that DF could become endemic to Athens, but the fact that most persons living in the metropolitan area got sick and became immune, combined with the enhanced mosquito control measures that would be initiated, led him to conclude that DF would not be a real threat during the following years. That same day, I Kathimerini claimed that effi cient measures against Stegomyia mosquitoes should suffi ce: covering any kind of water container and sewage-connected pipe or container and removing stagnant water from houses, kitchens, washing places, and fl ower pots.

1929-1931
Throughout the winter of 1928-29, sporadic articles citing DF, usually about unrelated matters, appeared in I Kathimerini. An often-recurring advertisement stated that many persons developed a hernia because of the "awful dengue disease," and true relief from it could be found only in the "incomparable invention" of a Mr. Em. Rousos. Additionally, a short advertisement said, "Against Dengue, Life Savers [candies] triumphed; they act the same way against the fl u." In May 1929, a series of articles in I Kathimerini described several fi ndings and conclusions of investigations led by the Hellenic Pasteur Institute. In addition to confi rmation of Stegomyia mosquitoes as the vector and lifelong immunity against DF, the scientists claimed that they also had produced an effi cient vaccine; however, it was not to be used commercially. The vaccine had been tested, according to the newspaper, on members of the leper colony on the small island of Spinalonga and on patients of a public psychiatric hospital near Athens.
On July 7, 1929, I Kathimerini cited Ministry of Health offi cials as saying that no new DF case had appeared in Athens; only 2 days later, I Kathimerini reported that a person suspected to have DF was brought to the hospital. On July 21, three presumed DF cases were reported from Thessaloniki, the second largest city in Greece, ≈300 km north of Athens. Isolated cases of diseases resembling DF were reported sporadically throughout Greece but were never confi rmed. Of those, the most important ones are ≈30 cases on the island of Syros in August 1931, of which a few supposedly were fatal. Yellow fever and an unknown spirochete infection were again discussed by the public (and discarded by health professionals); the unknown pathogen was declared to be transmitted by unidentifi ed mosquitoes. To positively diagnose the disease, 3 psychiatric patients in Athens were inoculated with infected blood. One of the 3 who had not acquired DF 3 years earlier became sick with signs and symptoms of DF; the other 2, who had acquired DF during the epidemic, remained healthy. (Relatives of the fi rst patient later sued the leader of the Hellenic Pasteur Institute team for attempted homicide. I Kathimerini did not report the outcome of this lawsuit.) Four months after these experiments, Institut Pasteur in Paris stated that the disease cases on Syros were caused by a spirochete.

Discussion
Reading the old news items sparks 3 issues. First, did the high number of fatal cases in 1928 result from dengue hemorrhagic fever? Two reports support this notion on the basis of the limited serologic study of Athenians born around the time of the epidemic (2,3). The data are consistent with a sequential infection in some of the population with dengue viruses of different serotypes, a conditio sine qua non for the appearance of dengue hemorrhagic fever (here DENV-1 and DENV-2). This notion was contradicted by another serologic analysis that indicated only DENV-1 was involved in the epidemic (11). According to statements by health offi cials, the newspaper articles alleged that no hemorrhagic illnesses were observed. However, this information contradicts the fi ndings from Aegina (10). Could the statement of the health offi cials have been false simply to calm the public?
Finally, several of the news items in 1928 cite physicians' assertions that DF patients from the previous year never became re-infected. This fact does not exclude the possibility of an asymptomatic or a misdiagnosed fi rst infection. The high death rate could have resulted from the incidence of 2 independent circumstances: highly virulent DENV-1 combined with the limited appearance of DENV-2, which could account for the hemorrhagic disease on Aegina and presumably elsewhere. The essence is that the question about the high death rate cannot be answered, and only a vast serologic study of the population, perhaps involving residents of the other areas in Greece where rates of death were high, might provide clues. However, 83 years after the epidemic, the pool of potential prospects continues to diminish.
The second issue regards the presence of Ae. aegypti mosquitoes in Greece. Except for infrequent sampling of specimens, the species seems to have disappeared from the country, as it has from most of the Mediterranean region (12,13). Two main reasons seem to account for its disappearance: 1) the widespread urban and rural antimalaria campaign based on insecticides and environmental management that lasted into the late 1950s (14); and 2) economic development that brought along measures and changes, such as running water, which helped reduce the number of containers in which mosquitoes reproduced. As in other places, malaria vectors, although found in much smaller numbers, have not disappeared from Greece and still can be identifi ed in entomologic collections (13). As in other European countries, Ae. albopictus mosquitoes are now found in several areas in Greece (15; J. Vontas, pers. comm.).
The third issue is, of course, whether such a devastating epidemic could recur. The social conditions, including medical sciences and health care, now differ substantially from the late 1920s, and recurrence of an epidemic of such an extent and magnitude is unlikely in Greece or elsewhere in Europe. On the basis of present standards, in the absence of Ae. aegypti mosquitoes, even a major Ae. albopictustransmitted DF epidemic is extremely unlikely, although minor events (in terms of numbers), such as the recent chikungunya epidemic in northern Italy (16), cannot be excluded. Recent cases of dengue (17,18) and chikungunya (19) in Europe resulting from international travel also support this notion (4).