Recent influenza activity in tropical Puerto Rico has become synchronized with mainland US

Abstract Background We used data from the Sentinel Enhanced Dengue Surveillance System (SEDSS) to describe influenza trends in southern Puerto Rico during 2012‐2018 and compare them to trends in the United States. Methods Patients with fever onset ≤ 7 days presenting were enrolled. Nasal/oropharyngeal swabs were tested for influenza A and B viruses by PCR. Virologic data were obtained from the US World Health Organization (WHO) Collaborating Laboratories System and the National Respiratory and Enteric Virus Surveillance System (NREVSS). We compared influenza A and B infections identified from SEDSS and WHO/NREVSS laboratories reported by US Department of Health and Human Services (HHS) region using time series decomposition methods, and analysed coherence of climate and influenza trends by region. Results Among 23,124 participants, 9% were positive for influenza A and 5% for influenza B. Influenza A and B viruses were identified year‐round, with no clear seasonal patterns from 2012 to 2015 and peaks in December‐January in 2016‐2017 and 2017‐2018 seasons. Influenza seasons in HHS regions were relatively synchronized in recent years with the seasons in Puerto Rico. We observed high coherence between absolute humidity and influenza A and B virus in HHS regions. In Puerto Rico, coherence was much lower in the early years but increased to similar levels to HHS regions by 2017‐2018. Conclusions Influenza seasons in Puerto Rico have recently become synchronized with seasons in US HHS regions. Current US recommendations are for everyone 6 months and older to receive influenza vaccination by the end of October seem appropriate for Puerto Rico.


| BACKG ROU N D
Influenza is a significant public health problem globally. Seasonal influenza has a high disease burden, reducing school attendance, increasing worker absenteeism and impacting daily productivity. 1 Since 2010, CDC estimates that influenza is associated with between 9.3-49 million illnesses, 140 000-960 000 hospitalizations and 12 000-79 000 deaths each year in the United States (US). 2 Though antivirals treat infection, vaccination remains the primary tool to prevent influenza-associated morbidity and mortality. 3 The US Advisory Committee on Immunization Practices (ACIP) recommends routine annual influenza vaccination for individuals, aged ≥6-months who do not have contraindications. 4 Influenza immunization programmes must consider disease seasonality to most effectively use immunization. 5  In high-income temperate countries, influenza has been well described. Most seasonal epidemics in temperate regions occur during the winter months, between November and March in the Northern Hemisphere and between April and September in the Southern Hemisphere. 6 These seasonal patterns are thought to be driven by annual changes in climate, contact rates, human immunity and other factors. 1,7,8 Some tropical and subtropical regions experience annual epidemics coinciding with local rainy seasons, 9 whereas others have semi-annual epidemics or yearround influenza activity without well-defined influenza seasons. 10,11 Influenza seasonality in tropical and subtropical regions is less understood. 9 Few studies examined influenza in Puerto Rico, and none  Figure S1). We also explored coherence between climate and influenza trends for Puerto Rico and HHS regions, to determine whether climatic differences explained differences in influenza trends.   Figure S1). 1

| Data analyses
We compared the number of confirmed influenza A and B cases in SEDSS to virologic surveillance data collected by US WHO and To explore if differences in seasonality between Puerto Rico and HHS regions were due to differences in climate, we obtained regional climate data (absolute humidity, temperature and precipitation) from 2012 to 2018 from the North American Regional Reanalysis data set from the National Oceanic and Atmospheric Administration (NOAA). 18 Daily climate data were temporally and spatially aggregated to weekly means for each HHS region and Puerto Rico.
Weekly mean time series data were normalized. In order to analyse the possible relationship to climate, we extended the timeframe considered for the seasonal component to 40-80 weeks to include potential climatic trends that might occur outside the typical seasonal periodicity. For each region, we used the Morlet wavelet decomposition to extract the major seasonal component (period: 40-80 weeks) for weekly influenza cases and the weekly mean absolute humidity, temperature and precipitation.

| Role of funding source
The US CDC funded the study, and participated in the study design, data analysis, data interpretation and preparation of the manuscript.
All authors had full access to study data, and all authors had final responsibility for the decision to submit for publication.  (Table 1). Overall, 14% of participants were positive for influenza; 9% for influenza A and 5% for influenza B viruses. Although there were no differences by gender, the percentage of participants with positive results for influenza A or B viruses differed by age (P < .001) and days after symptom onset (P < .001; Table 2).

| RE SULTS
We compared the time series and peak timing of overall influenza and type-specific patterns in Puerto Rico to the HHS regions.
In Puerto Rico (SEDSS), influenza peaks varied substantially be-   (Table S1). Similar coherence patterns are also observed when looking at influenza A and B.  To investigate possible climatic drivers of influenza A and B epidemics patterns within each geographic region, we assessed the coherence and phase differences between location-specific climate variables and type-specific influenza incidence (Figure 2A-I).
In terms of the overall time series climatic trends, Puerto Rico was consistently warmer and more humid over the time series compared with weekly average absolute humidity and temperature in HHS regions (Figure 2A 19,20 and epidemiological studies showed lower humidity is associated with the onset of influenza epidemics in the United

| D ISCUSS I ON
States. 21 Increased proximity between susceptible and infected hosts associated with cold weather is also frequently suggested as an important driver of influenza seasonality. 8

ACK N OWLED G EM ENTS
We acknowledge funding provided by the Centers for Disease Control and Prevention. We would like to thank the contribution of all study participants. We would also like to acknowledge the work of Olga Lorenzi and her support to this project.

CO N FLI C T O F I NTE R E S T
The authors report no conflict of interest.

AUTH O R CO NTR I B UTI O N S
GPB designed the study, contributed to the analyses, participated in the interpretation of results and wrote the first draft of the paper.

D I SCL A I M ER
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

S U PP O RTI N G I N FO R M ATI O N
Additional supporting information may be found online in the Supporting Information section.