Introduction

In March 2009, an outbreak of a novel influenza A (H1N1) pdm09 virus, now commonly referred to as the 2009 H1N1 pandemic influenza, was detected in Mexico and spread rapidly worldwide. The new virus spread so quickly that the World Health Organization (WHO) raised the pandemic level to category 6 in June 2009, which continued until August 2010 [1]. The WHO defined clinical advisory management guidelines to manage this pandemic, while European national authorities developed national programmes to face this situation, which included pregnant women [2–6]. Currently it is known that influenza A (H1N1) pdm09 infection can appear at different levels of severity as most pH1N1 infections have resulted in mild illness, similarly to seasonal influenza [7] in healthy persons, with a small proportion of more serious infection, severe illness, and even death. Most severe illnesses have been reported in persons with underlying medical conditions (diabetes, lung disease, cardiovascular disease and pregnancy) [8–10].

A high risk of influenza during pregnancy has been postulated. This increased risk can be related to the physiological changes of decreased pulmonary tidal volume and increased cardiac output. In addition, suppressed type 1 T-helper cell-mediated immunity impairs the maternal response to infection with several viruses, including influenza and smallpox [11]. To achieve successful term pregnancy, the mother’s immune system has to adapt to the developing embryo. This adaptation implies changes that can alter the mother’s susceptibility to some infections. For instance, it is known that pregnancy increases susceptibility to toxoplasmosis and lysteriosis, and that it also increases the severity of symptoms deriving from chickenpox and influenza infections [12]. This is supported by data from previous influenza epidemics and pandemics, which have shown that pregnant women present higher morbidity and mortality rates due to influenza infection than non-pregnant women [13–15]. The data collected between April 2009 and July 2009 from Australia and the USA suggest that pregnant women infected by the influenza A (H1N1) pdm09 virus were particularly vulnerable to morbidity and mortality [1, 16]. In the USA, a higher proportion of hospital admissions due to infection by the A (H1N1) pdm09 virus were observed for pregnant women than for the general population, and these women were also at higher risk of serious, or even deadly infection, by this virus [17, 18]. Furthermore, the health risks deriving from infection by a pandemic virus are even more relevant because these infections seemingly affect women of reproductive age more often and more severely than seasonal influenza epidemics [19].

Moreover, one major difference between the epidemiology of seasonal influenza and pH1N1 is the age distribution of influenza-related deaths. An estimated 90 % of deaths due to seasonal influenza complications occur in persons aged ≥65 years, with only 6 % occurring in persons aged 50–64 years, 3 % 5–49-year-olds, and <1 % in children of <5 years of age [20]. Yet during the pH1N1 pandemic, only 14 % of laboratory-confirmed influenza-related deaths occurred in persons ≥65 years of age, whereas 69 % were reported in adults aged 25–65 years [21], including pregnant women.

As compared with non-pregnant individuals, pregnant women are approximately 4–5 times more likely to develop severe disease [22]. The mortality rate of influenza A (H1N1) pdm09 virus infections varies substantially in different geographical regions [23, 24]. Susceptibility to influenza A (H1N1) pdm09 infection may vary in different races, as Wenger et al. [34] showed because native people of Alaska and Asian/Pacific Islanders were 2–4 times more likely to be infected by the (H1N1) pdm09 virus and hospitalised than white Caucasians. For these reasons, pregnant women should be considered a population for whom special prevention and treatment considerations for A (H1N1) pdm09 are necessary.

Very little is known about the characteristics of women with influenza A (H1N1) pdm09 and the factors affecting the recovery of infected patients in Spain.

The aim of this study was to investigate the main characteristics of non-vaccinated pregnant women who were hospitalised for influenza A (H1N1) pdm09 pandemic versus the pregnant women hospitalised for non-influenza-related reasons in Spain, and to also study the clinical features present in infected women for early diagnosis and future action programmes.

Materials and Methods

This comparative study forms part of a multi-centre case–control study of influenza A (H1N1) pdm09 carried out in seven Spanish Autonomous Communities (Catalonia, Andalusia, Madrid, the Basque Country, the Valencian Community, Castile-León, and Navarre) using a common protocol whose objective was to determine the impact of influenza A (H1N1) on the Spanish population [25].

Recruitment of Cases and Controls

This is a case–control study carried out in 28 hospitals in seven Spanish Autonomous Communities between 1 November 2009 and 30 June 2010 [25]. Of an initial study population of 813 persons hospitalised with laboratory-confirmed influenza A (H1N1) pdm09 infection and 2,972 persons attending the same hospitals, but without influenza-like illness, non-vaccinated pregnant women were selected. Fifty-one pregnant women with influenza aged between 15 and 44 years and 114 pregnant women without influenza aged 15–44 years were recruited for this study. A case was defined as a patient with a new diagnosis of influenza-like illness with clinical manifestations and a laboratory-confirmed influenza A (H1N1) pdm09 virus as assessed by a reverse transcriptase polymerase chain reaction (RT-PCR) assay [26], who were hospitalised for more than 24 h following the medical criteria for their illness. The patients with likely nosocomial infection (determined by onset of symptoms at 48 h post-hospitalisation) were excluded. Potential control individuals were excluded if the main reason for hospitalisation was respiratory illness. Informed consent was obtained from all the subjects prior to enrollment. The project was approved by the Ethics Committees of the participating hospitals. Further details of the study design have been provided elsewhere [25].

All the pregnant women, who were either cases or controls in the general study, were included in the present substudy. We specifically compared the pregnant women with influenza A (H1N1) pdm09 infection (cases) with the pregnant women without influenza who had been recruited from the same hospitals (controls). Women who had been vaccinated for pandemic influenza A (H1N1) pdm09 were excluded from the present study.

Data Collection

A face-to-face epidemiological enquiry was conducted, which was designed specifically for the multi-centre case–control study of influenza A (H1N1) pdm09 using a standardised protocol [25]. For the cases, the following variables were obtained from the personal interviews undertaken, along with the corresponding medical records from hospital databases: age, sex, country of birth, relevant medical conditions (pneumonia in the last 2 years, chronic obstructive pulmonary disease, asthma, other respiratory illness, chronic cardiovascular disease, arterial hypertension, cardiac insufficiency, nervous system, renal failure/nephritic syndrome, AIDS/HIV). The behavioural variables were ethnicity, level of education and smoking habit. The same variables were obtained from the uninfected controls, except the variables directly relating to the disease.

Statistical Analysis

Firstly, the characteristics between pregnant women with and without influenza A (H1N1) pdm09 were compared using the Chi square, Fisher’s exact and Student’s t test tests according to the nature of the variables. Subsequently, a logistic regression analysis was done to estimate the crude (cOR) and adjusted odds ratios (aOR) with 95 % confidence intervals (95 % CI) by taking into account any possible confounding variables (Wald’s test). Specifically, pre-pregnancy body mass index (BMI) was adjusted. The analyses were done with the SPSS software (version 17.0; Inc, Chicago III).

Results

The most relevant demographic characteristics of these pregnant women are summarised in Table 1, which compares the cases with and the controls without influenza A (H1N1) pdm09, and the corresponding ORs. Cases and controls did not differ in terms of age, pregnancy trimester, level of education, smoking habit, alcohol consumption (categorised: usual alcohol consumption or non-usual alcohol consumption) and diabetes status. However, obesity (BMI ≥ 30) was identified as a statistically significant risk factor for influenza infection in pregnant women (OR 3.03; 95 % CI 1.13–8.11). Higher, yet still non-significant, ORs were observed for Latin American pregnant women (aOR 3.71; 95 % CI 1.00–14.04).

Table 1 Demographic characteristics of cases with and controls without influenza A (H1N1) 2009, and the crude and adjusted ORs describing the associations between these characteristics and influenza infection in pregnant, non-vaccinated women in Spain
Full size table

The clinical characteristics of the cases with and the controls without infection, and the ORs describing the associations between these clinical characteristics and influenza infection in pregnant women, are summarised in Table 2. Cases and controls did not differ in terms of their medical histories of any comorbidities studied (pneumonia in the last 2 years, chronic obstructive pulmonary disease, asthma, other respiratory illness, chronic cardiovascular disease, arterial hypertension, cardiac insufficiency, nervous system, renal failure, nephritic syndrome, AIDS and HIV).

Table 2 Clinical characteristics of cases with and controls without influenza A (H1N1) pdm09 infection, and the ORs describing their associations with influenza infection in non-vaccinated pregnant women in Spain
Full size table

Table 3 reveals that the most prevalent symptoms in the influenza-infected pregnant women were temperature (mean 38.34 °C), cough (82.4 %), malaise (80.5 %), myalgia (56.1 %), and headaches (54.9 %).

Table 3 Prevalence of symptoms in cases of influenza A (H1N1) pdm09 virus infection in pregnant women in Spain
Full size table

Discussion

Summary of the Main Findings

The results obtained in the present study indicate a higher risk of the influenza A (H1N1) pdm09 disease in non-vaccinated pregnant women who are obese, and they also suggest an increased risk for subjects with obesity and for Latin American women living in Spain. No statistically significant differences for other characteristics features were observed, although a higher proportion of infection was noted for young pregnant women in pregnancy trimester 3, with a lower level of education and a higher incidence of toxic habits (alcohol consumption and smoking habit) than non-infected women.

The most frequent symptoms of influenza A (H1N1) pdm09 infection in pregnant women included a high temperature, cough, malaise, headache and myalgia, similarly to seasonal influenza, which can be differentiated by complementary microbiological tests.

Interpretation of the Results in Relation to Previous Findings

Our results support similar findings reported in previous studies, which reinforces the role of obesity as a risk factor for influenza A (H1N1) pdm09 in Spain. For instance, Van Kerkhove et al. [27–29] found that morbid obesity was a risk factor for ICU admission in patients with influenza A (H1N1) pdm09. Although the comparison made of average ages between pregnant women with and without pandemic influenza infection showed no statistically significant differences (Table 1), 27.5 % of infected pregnant women were younger than 20 years, whereas only 5.3 % of non-infected pregnant women fell in this age group. This result agrees with other analyses in which a higher risk for pandemic influenza virus was found [21, 30].

Basically, we found an increased risk of influenza A (H1N1) pdm09 virus during the third pregnancy trimester as compared to the first pregnancy trimester, although this risk was not statistically significant (probably due to the limited number of cases analysed). This result agrees with previous analyses [31] which reported an OR of 1.2 (95 % CI 1.1–1.4) in pregnancy trimester 3.

Our results also reveal that pregnant Latin American women are at an almost fourfold higher risk of being hospitalised for the influenza A (H1N1) pdm09 virus, suggesting that this influenza virus follows an ethnic pattern. Ethnic-dependent susceptibility to viral infection in the current pandemic has been discussed in previous research works undertaken in Turkey, New Zealand [32, 33], and people who are native from Alaska are more likely to be hospitalised native people [34]. The results presented herein also agree with the higher mortality described in South America [23]. To date, no explanation for these differences has been provided. One possible explanation for the low mortality rate recorded in China during the 2009 pandemic of influenza A (H1N1) pdm09 virus may be different ethnic susceptibility rather than strict strategies adopted to prevent infections [35]. Consequently, increased medical surveillance of this ethnic group, by means of centralised prevention and treatment efforts, might be advisable. An alternative hypothesis is that past contacts with similar viral strains might explain the greater resistance observed in the European population than in others, like more isolated populations in South America [23, 36] or Asia [35].

Previous AIDS or HIV infection, pathologies affecting the nervous and renal systems, chronic cardiovascular disease and pneumonia in the last 2 years are the main clinical characteristics associated factors to develop influenza A (H1N1) pdm09 infection in pregnant women. These results agree with previous findings [31, 37] obtained for increased risk of influenza A (H1N1) pdm09 infection in pregnant women with some of these comorbidities.

The most relevant symptoms found in pregnant women infected with the influenza A (H1N1) pdm09 virus include a high temperature, malaise, myalgia, cough and headache. These symptoms coincide with the typical symptoms of seasonal influenza in pregnant women, and it is not possible to clearly differentiate them. Our findings also suggest the importance of preventive measures, including immunity, which is also likely to benefit offspring [30].

Strengths and Limitations

This study compares pregnant women with and without the influenza A (H1N1) pdm09 virus. Most studies published on the influenza A (H1N1) pdm09 virus have compared pregnant and non-pregnant women [1, 16, 38], which means that this study is more specific. In addition, only non-vaccinated women were included in the analysis to prevent a confusion bias resulting from the protective effects of the vaccine.

We acknowledge several limitations in the present study, such as limited statistical power. We attempted to limit the information bias by following a stringent, standardised protocol and by training the interviewers. We believe that the case selection criteria have not changed with time, which limits the validity of the study because laboratory confirmations of infection by the pandemic strain of the influenza virus under study are always required. In order to minimise the confusion bias, we reviewed the most relevant literature and all the variables that can reasonably influence the results which, whenever necessary, were controlled in the logistic regression analysis. Furthermore, we believe that the study sample is representative of the target population because no participants differed in terms of urban/rural location or socio-economic status, and all the cases and controls were recruited according to the same criteria.

In conclusion, our study reveals a higher risk for influenza A (H1N1) pdm09 virus infection in pregnant, non-vaccinated women who are obese and Latin American. Specific efforts should be made to ensure that the risks of influenza during pregnancy are communicated to these women and to their healthcare providers in order to ensure that they can benefit from health care and influenza immunity.