Virus Etiology of Airway Illness in Elderly Adults.

Several new respiratory viruses have been discovered in humans since 2001: metapneumovirus (MPV), new coronavirus (CoV) strains NL63 and HKU1, bocavirus (BoV) 1, and rhinovirus (RV) species C.1, 2

Susceptibility to respiratory viral infections may be important especially in older age, but the viral etiology and clinical significance of respiratory illnesses in elderly adults is poorly documented.2, 3, 4 The aims of this study were to investigate the presence of viruses in elderly adults and to assess the association between viral infection and respiratory illness and between viral infection and chronic illness in individuals with an illness that requires hospitalization.

The study was conducted at Turku City Hospital between July 2007 and April 2009. Inclusion criteria included aged 65 and older and any disease necessitating hospitalization. Hospital episodes were divided in two groups depending on whether the individual had respiratory symptoms. The Ethics Committee of Turku University Hospital approved the study protocol.

Nasopharyngeal swab and serum samples were collected on hospital admission and after 2 weeks or at discharge from the hospital. Swab samples were analyzed for adenovirus, CoV NL63 and OC43, BoV, enteroviruses, MPV, RV, influenza A and B, PIV1–3, and Respiratory syncytial virus (RSV) A and B using polymerase chain reaction in the Department of Virology, Turku University Hospital. BoV infections were serologically confirmed in the Department of Virology, Helsinki University Hospital.

Seven hundred twenty‐nine swab samples were collected from 663 individuals. The mean age of individuals with and without respiratory symptoms was 83 ± 7. Asthma, chronic obstructive pulmonary disease (COPD), and rheumatic disease were more common in episodes with respiratory symptoms than in those without, and hypertension, stroke, dementia, and depression were less common (all P < .05).

Overall, 160 of 438 (37%) episodes with respiratory symptoms had more virus detections than 67 of 291 (23%) episodes without respiratory symptoms (P < .001, Table 1). Influenza (P = .006), CoV (P = .005), and MPV (P = .02) were detected more often when respiratory symptoms were present than when not. No acute HBoV1–4 infections occurred in the 396 episodes with respiratory symptoms studied using serology. Virus epidemics in elderly adults followed documented epidemics in the region. In episodes with respiratory symptoms two or more viruses were present in 22 (5%) samples, the most common viruses in coinfections being RV (59%) and influenza A virus (32%). The most common virus combination was RV with CoV, accounting for 25% of the coinfections. During episodes without respiratory symptoms, two or more viruses were detected in 14 (5%). The most common viruses were RSV and RV being present in 9 (64%) and 8 (57%) coinfections, respectively.

Table 1

Virus Findings in Hospital Care Episodes

Virus	Respiratory Symptoms
	Any, n = 438	With Dyspnea, n = 200	Without Dyspnea, n = 238	None, n = 291
Influenza virus	42 (10)a	20 (10)a	22 (9)a	12 (4)
A	34 (8)a	16 (8)a	18 (8)	11 (4)
B	10 (2)	5 (3)	5 (2)	2 (1)
Picornavirus	43 (10)	23 (12)	20 (8)	25 (9)
Rhinovirus	37 (8)	21 (11)	16 (7)	22 (8)
Enterovirus	8 (2)	3 (2)	5 (2)	3 (1)
Parainfluenza virus	33 (8)	13 (7)	20 (8)	16 (6)
Type 1	7 (2)a	3 (2)	4 (2)a	0 (0)
Type 2	4 (1)	0 (0)	4 (2)	3 (1)
Type 3	22 (5)	10 (5)	12 (5)	13 (4)
Coronavirus	29 (7)a	11 (6)a	18 (8)a	6 (2)
NL63	14 (3)	7 (4)	7 (3)	3 (1)
OC43	16 (4)a	5 (3)	11 (5)a	3 (1)
Respiratory syncytial virus	26 (6)	14 (7)	12 (5)	21 (7)
A	7 (2)	4 (2)	3 (1)	8 (3)
B	16 (4)	8 (4)	8 (3)	11 (4)
Metapneumovirus	8 (2)a	5 (3)a	3 (1)	0 (0)
Bocavirus 1	2/389 (1)	1/176 (1)	1/213 (0)	0/289 (0)
Adenovirus	2 (0)	2 (1)	0 (0)	2 (1)
Multiple viruses
≥1	160 (37)a	76 (38)a	88 (37)a	67 (23)
≥2	22 (5)	12 (6)	10 (4)	14 (5)
≥3	7 (2)	5 (3)	2 (1)	2 (1)
≥4	1 (0)	1 (1)	0 (0)	0 (0)

Chi‐squared test and Fischer exact test (when counts <5) were used.

Differed (P < .05) from episodes with no respiratory symptoms.

In the 729 episodes, age (range 65–100) was not associated with detection of respiratory viruses (P = .49), but the presence of a virus, especially influenza (odds ratio (OR) = 1.02, 95% confidence interval (CI) = 0.01–1.04) and parainfluenza viruses in the upper airways (OR = 1.02, 95% CI = 1.00–1.03), was positively associated with the individual's weight.

A virus was detected in 23% of episodes without respiratory symptoms, which is consistent with findings in individuals of all ages.5 A striking difference between these elderly adults and children was the low number of BoV and the high number of RSV coinfections in subjects without respiratory symptoms.

The detection rate for the newly discovered BoV 1 was low, possibly because of short‐term local replication or mere mucosal contamination of virus from grandchildren; there were no genuine acute BoV 1 infections. This is consistent with previous studies of adults;6, 7 BoV 1 respiratory infection seems to be more a pediatric problem.8 No studies have investigated BoV infections in individuals with a mean age of 80 and older.

An association has been reported between body mass index and severe influenza‐like illness,9 although only in people younger than 60, and the viruses were not identified. Obesity may impair vaccine‐induced immunity and make obese individuals more susceptible to influenza.10 The current study shows that there is an association between respiratory virus detection and weight in elderly adults.

In conclusion, influenza virus and RV were the most detected viruses in episodes with respiratory symptoms, but overall, symptomatic and asymptomatic respiratory virus infections are relatively uncommon in elderly adults. Weight was associated with virus detection. The detection rates for the newly discovered BoV 1–4 were low in elderly adults.