European Journal of Clinical Microbiology and Infectious Diseases

, Volume 23, Issue 6, pp 425–433

Dengue: an arthropod-borne disease of global importance


    • Department of Internal MedicineSlotervaart Hospital
  • J. Wagenaar
    • Department of Internal MedicineSlotervaart Hospital
  • D. P. M. Brandjes
    • Department of Internal MedicineSlotervaart Hospital
  • E. C. M. van Gorp
    • Department of Internal MedicineSlotervaart Hospital

DOI: 10.1007/s10096-004-1145-1

Cite this article as:
Mairuhu, A.T.A., Wagenaar, J., Brandjes, D.P.M. et al. Eur J Clin Microbiol Infect Dis (2004) 23: 425. doi:10.1007/s10096-004-1145-1


Dengue viruses cause a variable spectrum of disease that ranges from an undifferentiated fever to dengue fever to the potentially fatal dengue shock syndrome. Due to the increased incidence and geographical distribution of dengue in the last 50 years, dengue is becoming increasingly recognised as one of the world’s major infectious diseases. This article will review clinical and diagnostic aspects of dengue virus infections. It also presents our current knowledge of the pathophysiology of severe dengue and addresses the importance of dengue virus infections in those travelling to parts of the world where dengue is endemic.


The last 50 years witnessed a resurgence of dengue fever epidemics and an emergence of dengue haemorrhagic fever (DHF) and dengue shock syndrome (DSS) throughout tropical and subtropical regions around the world [1]. The causative agents of these syndromes, dengue viruses, are members of the Flaviviridae family and occur as four antigenically related but distinct serotypes, designated DEN-1, DEN-2, DEN-3, and DEN-4. The viruses share many characteristics with other flaviviruses, having a single-stranded RNA genome surrounded by an icosahedral scaffold and covered by a lipid envelope. The complete virion is 50 nm in diameter and contains an 11-kb plus-sensed RNA genome that is composed of seven nonstructural (NS) protein genes and three structural protein genes: core (C, 100 amino acids), membrane (M, 75 amino acids), and envelope (E, 595 amino acids) [2, 3]. The domains responsible for neutralization, fusion, and interactions with virus receptors are associated with the envelope protein. The order of proteins encoded is 5′-C-prM(M)-E-NS1-NS2A-NS2B-NS3-NS4A-NS4B-NS5-3′ [3].

Dengue viruses are transmitted to humans by the bite of infective female mosquitoes of the genus Aedes. The most efficient epidemic vector is Aedes aegypti, although Aedes albopictus and Aedes polynesienses are also both involved in dengue outbreaks. Several factors have been implicated in the global resurgence of dengue: failure to control the Aedes populations, increased airplane travel to dengue-endemic areas, uncontrolled urbanisation, and an unprecedented population growth [4]. With the spread of A. albopictus to Europe as well as the climate changes that have occurred there, conditions are becoming conducive to further dengue transmission and to a major European outbreak of dengue [58].


The first epidemics of dengue-like disease date back to 1779–1780, when outbreaks occurred in Batavia (Jakarta), Cairo, and Philadelphia [3, 9]. The near simultaneous occurrence of dengue outbreaks on three continents indicates that dengue viruses and their mosquito vector have had a worldwide distribution in the tropics for more than 200 years. Until the Second World War, epidemics occurred on almost all continents every 10–30 years, mainly because both the viruses and the mosquito vector relied on sailing vessels for transport from countries of the tropics [3, 9]. During most of this time, dengue was a benign, nonfatal disease that was characterised by high-grade fever and severe bone and back pain. After 1945, dengue cases complicated by haemorrhage and shock (DHF and DSS) were increasingly documented in Southeast Asia. The first reported DHF epidemic occurred in the Philippine Islands in 1953–1954; after that, epidemic activity of DHF intensified in that region, although it remained confined to Southeast Asia through the 1970s [10]. In the 1980s and 1990s, however, epidemic DHF spread west into India, Pakistan, Sri Lanka, and the Maldives and east into China.

In the Americas, dengue was not considered a major public health problem until recent years. Attempts to eradicate A. aegypti from the Americas were undertaken in the 1950s. Although this program, coordinated by the Pan American Health Organisation, was successful in several countries, it failed to eradicate A. aegypti from the whole region. After support for mosquito surveillance and control programs was reduced, most countries became reinfested with A. aegypti by the end of the 1970s. This, along with the emergence of new dengue serotypes, resulted in major epidemics throughout the region and the emergence of DHF. In 1981, the first major DHF epidemic in the Americas occurred in Cuba. During that outbreak, an estimated 344,000 dengue infections occurred and over 116,000 hospitalised patients and 10,000 cases of DHF were reported from May to October [11]. Since the reinfestation of Central and South America, epidemics caused by multiple serotypes (hyperendemicity) are more frequent and the geographic distribution of dengue viruses and their mosquito vectors has expanded. Moreover, by 1997, DHF had emerged as a disease entity in several major and many minor epidemics in tropical and subtropical countries of the Americas (Fig. 1). Today an estimated 50–100 million cases of dengue fever and 500,000 cases of DHF, resulting in around 24,000 deaths, occur annually, depending on the epidemic activity [4, 12]. Over half the world’s population lives in areas potentially at risk for dengue transmission, making dengue the most important human viral disease transmitted by arthropod vectors in terms of morbidity and mortality.
Fig. 1

Countries/areas reporting dengue cases to the World Health Organisation in 2001

Clinical features

Each of the four dengue virus serotypes can cause a spectrum of disease ranging from a clinically inapparent infection to a potentially lethal disease. The World Health Organisation has defined strict criteria for the classification of dengue, based on observations made during epidemics that primarily occurred in Thailand [12]. In addition to asymptomatic infections, the following grades are recognised according to the severity of disease: nonspecific febrile illness, classic dengue fever, DHF (grade I and II), and DSS (grade III and IV) (Fig. 2).
Fig. 2

Classification of symptomatic dengue virus infections [12]

Dengue fever

The incubation period after the mosquito bite is 3–8 days. Infants and young children often have a nonspecific febrile illness that can hardly be differentiated from other viral illnesses. The more severe cases of dengue fever are usually seen in older children and are characterised by a rapidly rising temperature (≥39°C) that lasts 5–6 days, sometimes returning to almost normal in the middle of the febrile period (biphasic or saddle-back temperature curve). The febrile period is accompanied by severe headache, retro-orbital pain, myalgia, arthralgia, nausea, and vomiting. Over half of infected people report a rash during the febrile period that is initially macular or maculopapular and becomes diffusely erythematous, sparing small areas of normal skin (“islands of white in a sea of red”) [9, 13]. Minor haemorrhagic manifestations like petechiae, epistaxis, and gingival bleeding do occur. Severe haemorrhage is unusual. Dengue fever may be very incapacitating, but its prognosis is favourable and recovery generally occurs after 7–10 days of illness.

Dengue haemorrhagic fever and dengue shock syndrome

DHF is defined as an acute febrile illness with haemorrhagic manifestations, thrombocytopenia (≤100,000 cells/mm3), and evidence of an increased vascular permeability resulting in loss of plasma from the vascular compartment. Hypoproteinaemia, an elevated haematocrit, and evidence of serous effusion are distinctive indicators of plasma leakage. When plasma loss becomes critical, it may result in DSS. The World Health Organisation defines DSS as DHF with circulatory failure as manifested by a rapid, weak pulse with narrowing of the pulse pressure (≤20 mmHg, regardless of pressure levels, e.g. 100/90 mmHg) or hypotension with cold, clammy skin and restlessness [12]. In Asia, DHF and DSS mainly affect children under 15 years of age in hyperendemic areas [14, 15]. The age distribution is different in the Americas, where these syndromes occur in all age groups. However, the majority of fatalities during epidemics in the Americas occur in children [1618].

The incubation period for DHF and DSS is similar to that of dengue fever. Patients with DHF or DSS initially present with symptoms resembling dengue fever, including a high temperature (≥39°C), arthralgia, myalgia, headache, and vomiting [12]. Haemorrhagic manifestations usually appear after 3–4 days and may vary from a positive tourniquet test and petechiae to haemorrhage from the gastrointestinal tract, nose, and gums. The critical stage is reached after 3–7 days of fever. Approximately 24 h before and 24 h after fever defervescence, signs of circulatory failure of varying severity may appear. Several symptoms and signs occur before defervescence and may serve as warning signs that DHF and DSS are impending: generalised abdominal pain, persistent vomiting, change in the level of consciousness, a sudden drop in the platelet count, and a rapid rise in the haematocrit. By this time, most patients also have evidence of pleural effusion on the chest radiograph [19]. If plasma loss continues and becomes excessive, the patient’s situation can progress into profound shock. The outcome of DHF and DSS depends largely on early diagnosis and the immediate replacement of fluid. Uncorrected shock can give rise to the development of metabolic and electrolyte disturbances often complicated by severe bleeding from the gastrointestinal tract and other organs. Once shock is overcome, survivors usually recover within 2–3 days, although pleural effusion and ascites may still be present.

Other manifestations

An increasing number of dengue infections have been related to other unusual manifestations. These include dengue fever with severe haemorrhage, fulminant liver failure, cardiomyopathy, and neurological phenomena such as altered consciousness, convulsions, and coma resulting from encephalitis and encephalopathy [4, 12, 20, 21]. Previously, neurological manifestations were ascribed to nonspecific complications secondary to DHF and DSS. Possible causes of dengue encephalopathy include hypotension, cerebral oedema, focal haemorrhage, hyponatraemia, and fulminant hepatic failure. However, a recently documented possibility is the invasion of the central nervous system [20]. Other unusual presentations include ocular manifestations [22, 23].

Pathology and pathogenesis of dengue haemorrhagic fever and dengue shock syndrome


On pathological examination of fatal cases, serous effusion in the pleural and abdominal cavities and haemorrhage in the skin, subcutaneous tissue, gastrointestinal tract, heart, and other organs are almost invariably present [3, 12]. Histopathological changes are observed in three major organ systems: the liver, lymphocytic tissue, and the vascular system [24]. Most changes in the liver are focal and mild and are similar to those observed in the early phase of yellow fever infection: fatty metamorphosis, the formation of Councilman bodies, and degeneration of liver cells and Kupffer cells [25]. The recovery of dengue virus from the liver of children with fatal dengue and the detection of dengue viral antigens in liver cells suggest that some of these changes may result from a direct viral infection of the liver [2628]. Changes secondary to circulatory failure, such as perisinusoidal oedema, congestion, and haemorrhage, may also be present. Lymphocytic tissue in the spleen, lymph nodes, and liver shows increased activity as evidenced by both proliferation of lymphoblastoid cells and lymphocytic phagocytosis, indicating considerable turnover of lymphocytes [25]. Upon the finding of exudates in the extravascular compartment and serosal cavities along with haemorrhage, one would expect an endothelium severely damaged to the extent of cellular necrosis. However, light and electron microscopy of blood vessels do not show such changes in the vascular wall. Rather, a swelling of endothelial cells and diapedesis of erythrocytes through vessel walls with perivascular infiltration by lymphocytes and mononuclear cells may be observed [25, 29].


The absence of severe pathology in fatal cases and the fact that most surviving patients recover rapidly suggest that DHF and DSS are secondary to the action of biologic mediators that are capable of producing severe illness with minimal structural injury. Among possible mediators that cause increased vascular leakage, complement activation fragments C3a and C5a, both anaphylatoxins, and vasoactive cytokines, such as interferon-γ and TNF-α, have been implicated [30, 31]. Vasculopathy, thrombocytopenia, and thrombocytopathy are thought to give rise to the characteristic haemorrhagic phenomena. Increasing evidence, however, suggest defects in coagulation and fibrinolysis play a pivotal role in the development of severe disease [32]. Many studies have attempted to unravel the mechanism underlying the deterioration of dengue virus infections into DHF and DSS. Despite these efforts, the mediators that increase vascular permeability and the precise mechanism of the haemorrhagic phenomena are still a controversial matter and remain to be elucidated.

The simultaneous occurrence of several risk factors related to individual, epidemiological, and viral aspects determines whether a person in a given population will present with the clinical picture of DHF or DSS and whether an epidemic follows [1]. Almost all patients with DHF or DSS have had a previous infection with at least one of the four serotypes of dengue viruses [3337]. Results from numerous studies suggest that during a secondary infection with a different serotype, the presence of cross-reactive, non-neutralising antibodies enhances the efficiency with which dengue virus infects susceptible cells. This can lead to an activation of cross-reactive dengue-virus-specific T lymphocytes from the primary infection and a release of a number of cytokines and chemical mediators [30, 38]. This phenomenon, called antibody-dependent enhancement (ADE), may explain why infants under 1 year of age sometimes experience a more severe form of dengue due to passively acquired heterotypic maternal dengue antibodies [39]. However, DHF and DSS have also been reported in patients with primary infections [4042].

Other individual risk factors, like age, sex, nutritional status, and genetic predisposition, have been correlated with the development of severe disease [35, 43, 44]. It has been suggested that microvascular permeability and fragility in children is considerably greater than in adults [45]. The suggestion that certain persons or groups are more susceptible or resistant to severe dengue is derived from epidemiological studies. In the 1981 dengue outbreak in Cuba, it was observed that black individuals were relatively resistant to DHF and DSS, and the existence of a “resistance gene” was speculated [46]. Several genes have been found to be associated with the varying degrees of clinical illness following exposure to dengue virus [4749]. These genes could explain at least in part why some individuals develop severe disease whereas others experience a nonspecific fever or no symptoms at all. Viral risk factors include infection with a more virulent dengue virus and the size of the viral load. Several studies found that DEN-2 caused more severe disease in hospitalised patients [3337]. Dengue severity and outcome were also correlated with peak virus titres early in the course of disease and during defervescence [33, 50].


Laboratory diagnosis of dengue relies on the recovery of the virus by culture or on the detection of antibodies to the dengue virus [12]. Although culture is the definitive diagnostic test, several practical considerations limit its use. First, virus isolation is mainly successful when attempted using clinical samples obtained during the acute phase of disease. Already within a day or two after fever defervescence, antibodies to the dengue virus will begin to interfere with virus culture. Second, samples obtained for the detection of dengue virus by culture require proper handling. If storage and transportation conditions are not optimal, the virus will be inactivated in the samples. And third, not all laboratories are capable of culturing dengue viruses due to financial and safety reasons [12]. Several promising laboratory techniques have been introduced in recent years with which either viral RNA or viral antigen can be detected in samples obtained from infected patients. Amplification of dengue RNA by reverse transcriptase polymerase chain reaction (RT-PCR) has proved successful [5153]. Dengue RNA may be detectable during convalescence, when virus culture would fail due to the presence of dengue antibodies. Using RT-PCR, dengue RNA in human samples can be quantified [54]. Nevertheless, this approach remains a problem as it is difficult to use on a large scale and is still considered investigational [12]. Recently, it was found that the nonstructural protein NS1 was circulating during the acute phase of infection [55]. Capture enzyme-linked immunosorbent assays (ELISAs) have been developed to detect the NS1 protein, and several reports indicate that detection of the NS1 protein may allow early diagnosis of infection [5557]. Although an antigen detection kit is commercially available, its potential diagnostic use still needs to be evaluated properly.

Serological assays are most widely used in routine practice to confirm dengue virus infections and to differentiate between a primary and a secondary infection [12]. Primary infections are characterised by an increase in the levels of dengue-virus-specific immunoglobulin M (IgM) antibody 3–5 days after the onset of fever. IgM antibody titres continue to rise for 1–3 weeks and remain detectable for up to 6 months. Dengue-virus-specific immunoglobulin G (IgG) antibodies increase shortly after the initial rise in IgM antibodies to a modest degree and remain detectable for life [12, 58]. In secondary infections, the IgM antibody level is generally lower than in primary infections, while IgG antibody levels rise rapidly from 1 to 2 days after the onset of symptoms [12, 58]. The most commonly used serological techniques for the diagnosis of dengue infection are an ELISA that detects IgG or IgM antibodies and the haemagglutination-inhibition test. Traditionally, the haemagglutination-inhibition test has been used for the diagnosis of dengue and requires paired acute and convalescent sera collected 1 week or more apart for definitive diagnosis. However, this test requires samples to be pretreated, and the variability in methods used in different laboratories has compromised its general applicability. Today, many laboratories rely on other immunosystems for the diagnosis of dengue infections: ELISAs, immunochromatographic assays, or dot-blot assays [5963]. The most widely used are an IgM antibody-capture enzyme-linked immunosorbent assay (MAC-ELISA) or a combination of IgM-capture and IgG-capture ELISAs. The interpretation of serological test results may be complicated by the occurrence of cross-reactive antibodies to antigenic determinants shared by other members of the flavivirus family (i.e. Japanese encephalitis virus). Commercial kits for the rapid detection of IgG as well as IgM antibodies have become available and seem to be useful and reliable for serodiagnosis of dengue virus infection [64].

Prevention and treatment


Although urgently needed, no vaccine is currently available to protect against dengue virus infections. An effective vaccine will have to provide protective immunity against all four dengue viruses, bearing in mind concerns about ADE enhancement [65]. Promising results from studies with a live attenuated chimeric vaccine against Japanese encephalitis and the development of new dengue vaccines using a similar methodology provide ample basis for optimism that a safe vaccine is feasible [6668]. Until a vaccine is available, large dengue outbreaks can only be stopped by controlling vector populations. In addition to insecticide application, breeding sites of Aedes mosquitoes should be reduced vigorously or properly managed. Stagnant water should be avoided and containers of standing water should be emptied. Personal protection measures, such as the use of mosquito repellents, protective clothing, or insecticides, are necessary to protect travellers and residents of endemic areas against dengue virus infections [12]. Effective and sustainable prevention of dengue outbreaks must include active community participation, a good public health infrastructure, and political will. Although eradication programs may impose a financial burden on developing countries, the failure of these programs in the Americas is a good example of what can happen when political and community support is reduced. Recent education programmes have resulted in an increased awareness of the seriousness of severe dengue virus infections among residents living in dengue-endemic areas. However, such programmes did not always bring about behavioural changes or result in adequate control of mosquito habitats [69, 70].


Currently no specific therapeutic agent exists for dengue. Mild dengue infections may be treated with oral hydration and antipyretics only. Paracetamol should be used instead of aspirin or other nonsteroidal anti-inflammatory drugs to avoid the increased risk of Reye’s syndrome and haemorrhage [71, 72]. Patients need to be monitored closely for signs of shock until at least 24 h after defervescence. For patients suffering from DSS, the mainstay of therapy is early and effective replacement of plasma loss. The World Health Organisation recommends immediate volume replacement with Ringer’s lactate, Ringer’s acetate, or 5% glucose diluted in physiological saline, followed by plasma or colloid solutions in the event that shock persists [12]. Recently, two randomised controlled trials evaluated therapeutic responses to colloid and crystalloid solutions [73, 74]. Results indicate that Ringer’s lactate performed the least well and that the more severely ill patients, identified by a narrow pulse pressure (≤10 mmHg), would benefit more from initial resuscitation with colloid solution than with crystalloid solution. However, these studies were underpowered and the findings will need to be confirmed in large studies in which patients are stratified by severity of disease on admission.

Since volume replacement therapy alone may be insufficient for those admitted late in the course of disease, additional interventions for the treatment of dengue virus infections have been sought. Carbozochrome sodium sulfonate (AC-17), a drug that blocked capillary permeability in an animal model, failed to prevent dengue vascular permeability or shock in humans [75]. High-dose methylprednisolone did not reduce mortality in severe DSS, although prolonged thrombocytopaenia following DHF responded well to corticosteroids [76, 77]. In addition, several case reports described the potential use of desmopressin, high dosages of immunoglobulin, and recombinant activated factor VII in dengue virus infections [7880]. The administration of these drugs was followed by an immediate clinical response, but clinical trials evaluating their effectiveness are still awaited.

Dengue in travellers to the tropics

The number of people travelling to dengue-endemic countries is rising. The result is not only that dengue infections among travellers are encountered more often as a febrile illness, but also that these travellers may serve as vehicles for further spread [81, 82]. Imported dengue virus infections have been reported from several nonendemic countries, and data suggest that infection with one of the dengue viruses is one of the most frequent causes of febrile illness among tourists and people working in dengue-endemic areas [8390]. In particular, young adults who travel to Asia for long periods of time are at increased risk [9193]. Although reports like these are relatively frequent, they do not allow for an estimation of the true incidence of dengue virus infections among travellers.

Frequency of occurrence of dengue virus infections in long-term and short-term travellers

Only a small number of studies prospectively studied the incidence and risk of dengue virus infections among travellers (Table 1). The risk of infection with dengue virus was assessed in 204 North American relief workers visiting Puerto Rico after Hurricane George in 1998. Eleven of 12 participants reporting a dengue-like illness provided follow-up blood specimens for laboratory confirmation. None of these participants had laboratory evidence of dengue virus infection [94]. This is in contrast to what was observed in a large cohort of Israeli travellers with free access to consultations provided by Israeli physicians in five Southeast Asian countries (Thailand, Nepal, India, Vietnam, and China) in 1998. Dengue fever was diagnosed and laboratory confirmed in 17 travellers [95]. The authors state that the calculated attack rate of 3.4 infections per 1,000 persons is probably an underestimation because some patients recovered without seeking medical attention and others may not have undergone serological testing at all. Pretravel prevention advisories probably contributed to protective behaviour in the first study and may explain the observed differences. Both studies focussed only on those individuals who reported a dengue-like illness and did not evaluate the number of persons who seroconverted.
Table 1

Incidence of dengue virus infections and dengue-like illnesses in travellers to endemic areas

Type of traveller



Duration of travel

No. of travellers

No. (%) of seroconverters


Israeli long-term travellers [93]

Not stated

Southeast Asia, South America, and Africa

At least 3 months


7 (6.3)

Only 4 (3.8%) travellers of the group seroconverted; travellers recalled a febrile illness during travel

Dutch travellers aged 15 years and older [96]



1–13 weeks


13 (2.9)

Dengue-like illness reported in three of these 13 (0.7%) travellers

Israeli travellers with free access to medical services [95]


Southeast Asian countries

Not stated


Not evaluated

17 (0.3%) travellers contracted laboratory-confirmed dengue in Thailand

North American relief workers [94]


Puerto Rico

Average stay of 16 days


Not evaluated

12 (5.9%) travellers reported dengue-like illnesses, but diagnosis was not laboratory confirmed

In a prospective study among 447 Dutch short-term travellers to Asia, a probable dengue infection was found in 13 (2.9%) travellers, but only 3 of these 13 travellers reported experiencing a dengue-like illness during their stay abroad [96]. Dengue seroconversions also occurred in 7 of 104 Israeli young adults who travelled to tropical countries. Only four travellers reported being ill while abroad [93]. The discrepancy between infection and clinical expression are comparable with what is observed in areas of endemicity, where 14–87% of infections may go unnoticed [92].


Since the 1950s, the resurgence of dengue fever and emergence of severe forms of dengue cause significant public health problems in tropical parts of the world. Factors contributing to the increased incidence of dengue are intensifying. One of the many factors includes increased air travel, which makes those travelling to dengue-endemic countries not only at increased risk for infection but also makes them a potential vector for further spread of the disease. Numerous reports have demonstrated that dengue should be considered in the differential diagnosis of febrile travellers, particularly upon their return from destinations in the South Pacific. Healthcare providers should therefore have an understanding of the infection’s spectrum, its diagnosis, and its appropriate treatment. The incomplete understanding of the pathogenesis of dengue has delayed the development of an effective vaccine and specific therapeutic agents. Until the Aedes mosquito can be effectively controlled or a cost-effective vaccine developed, the occurrence of dengue can be expected to continue to escalate not only in (sub-)tropical parts of the world but also in dengue-free countries that harbour the Aedes mosquito.


The Royal Netherlands Academy of Arts and Sciences funded a project on Dengue virus infections (project number 99-MED-04) in which all authors are involved. Funding sources had no involvement in the writing of this report.

Copyright information

© Springer-Verlag 2004