Skip to main content
. 2017 Oct 16;66(6):865–876. doi: 10.1093/cid/cix903

Table 2.

Risk Factors for Laboratory-confirmed Influenza Illness Among Cohort Participants

Characteristic Incidence Rate per 1000 PY of Follow-up (95% CI) Univariable Analysisa Multivariable Analysisa,b
IRR (95% CI) P Value IRR (95% CI) P Value
Sex
 Male 31.4 (16.9–58.3) 1 1
 Female 34.6 (22.1–54.2) 1.21 (0.52–2.84) .66 0.88 (0.40–1.93) .74
Age group, y
 18–29 20.9 (8.7–50.2) 1 1
 30–39 39.9 (23.2–68.7) 1.73 (0.60–4.97) 1.55 (0.54–4.43)
 ≥40 36.3 (20.1–65.6) 1.33 (0.43–4.05) .30 1.42 (0.47–4.28) .70
HIV status
 Uninfected 14.5 (6.0–34.7) 1 1
 Infected 46.0 (30.8–68.6) 3.21 (1.22–8.41) .02 2.75 (1.02–7.44) .03
Medical history
 Previous pulmonary tuberculosis 28.2 (23.3–51.1) 0.82 (0.28–2.43) .87
 Pneumonia in past 5 y 54.3 (24.4–120.9) 1.86 (0.75–4.57) .14
 Body mass index <18.5 kg/m2 40.3 (13.0–125.0) 0.81 (0.25–2.68) .73
Housing characteristics
 No. of children aged <5 y in household
  0 36.2 (22.8–57.5) 1
  1 27.8 (13.9–55.6) 0.77 (0.33–1.77)
  ≥2 37.3 (12.0–115.5) 1.03 (0.30–3.50) .80
 No. of individuals aged ≥5 y in household
  0–2 25.5 (10.6–61.3) 1
  3–4 28.1 (15.1–52.3) 1.10 (0.38–3.22)
  ≥5 44.7 (26.5–75.4) 1.76 (0.63–4.87) .41
 Crowding indexc
  <1.5 16.3 (6.12–43.5) 1 1
  1.5–2.4 47.1 (28.8–76.8) 2.88 (0.96–8.62) 3.41 (1.12–10.36)
  ≥2.5 32.0 (16.6–61.5) 1.95 (0.60–6.35) .11 2.06 (0.62–6.83) .06
Socioeconomic characteristics
 Highest level of education
  Never attended/primary 47.6 (30.0–75.6) 1
  Secondary/tertiary 22.5 (12.5–40.7) 0.47 (0.22–1.00) .05
 Employment
  No 49.6 (28.8–85.4) 1
  Yes 26.4 (16.2–43.1) 0.53 (0.26–1.11) .09
 Food security: difficulty obtaining food
  Never 25.9 (13.9–48.2) 1 1
  1–2 times/mo 20.7 (9.9–43.5) 0.80 (0.30–2.10) 0.71 (0.27–1.88)
  >2 times/mo 87.2 (49.5–153.6) 3.35 (1.45–7.76) .005 3.09 (1.30–7.36) .006

Abbreviations: CI, confidence interval; HIV, human immunodeficiency virus; IRR, incidence rate ratio; PY, person-years.

aIRRs estimated for the incidence of laboratory-confirmed influenza, using Poisson regression.

bAdjusted for sex, age group, HIV status, crowding index, and food security.

cThe crowding index was calculated as the number of persons in the household divided by the number of sleeping rooms.