A s
tudy of personal, indoor, and ou
tdoor exposure
toPM
2.5 and associa
ted elemen
ts has been carried ou
t for37 residen
ts of
the Research Triangle Park area in Nor
thCarolina. Par
ticipan
ts were selec
ted from persons expec
ted
to be a
t eleva
ted risk from exposure
to par
ticles, andincluded 29 persons wi
th hyper
tension and 8 cardiac pa
tien
tswi
th implan
ted defibrilla
tors. Par
ticipan
ts were moni
toredfor 7 consecu
tive days in each of four seasons. Onegoal of
the s
tudy was
to es
tima
te
the con
tribu
tion of ou
tdoorPM
2.5 to indoor concen
tra
tions. This depends on
theinfil
tra
tion fac
tor
Finf,
the frac
tion of ou
tdoor PM
2.5 remainingairborne af
ter pene
tra
ting indoors. Af
ter confirming wi
thour measuremen
ts
the findings of previous s
tudies
tha
t sulfurhas few indoor sources, we es
tima
ted an average
Finffor each house based on indoor/ou
tdoor sulfur ra
tios. Thesees
tima
tes ranged from 0.26
to 0.87, wi
th a median valueof 0.55. Since
these es
tima
tes apply only
to par
ticles of sizesimilar
to
tha
t of sulfur par
ticles (0.06-0.5
ti
ties/mgr.gif">m diame
ter),and since larger par
ticles (0.5-2.5
ti
ties/mgr.gif">m) have lower pene
tra
tionra
tes and higher deposi
tion ra
tes,
these es
tima
tes arelikely
to be higher
than
the
true infil
tra
tion fac
tors for PM
2.5as a whole. In summer when air condi
tioners were inuse,
the sulfur-based infil
tra
tion fac
tor was a
t i
ts lowes
t(averaging 0.50) for mos
t homes, whereas
the average
Finffor
the o
ther
three seasons was 0.62-0.63. Using
thedaily es
tima
ted infil
tra
tion fac
tor for each house, we calcula
ted
the con
tribu
tion of ou
tdoor PM
2.5 to indoor air concen
tra
tions.The indoor-genera
ted con
tribu
tions
to indoor PM
2.5 hada wider range (0-33
ti
ties/mgr.gif">g/m
3)
than
the ou
tdoor con
tribu
tions(5-22
ti
ties/mgr.gif">g/m
3). However, ou
tdoor con
tribu
tions exceeded
the indoor-genera
ted con
tribu
tions in 27 of 36 homes. Asecond goal of
the s
tudy was
to de
termine
the con
tribu
tionof ou
tdoor par
ticles
to personal exposure. This is de
terminedby
the "ou
tdoor exposure fac
tor"
Fpex,
the frac
tion ofou
tdoor PM
2.5 con
tribu
ting
to personal exposure. As wi
th
Finf, we es
tima
ted
Fpex by
the personal/ou
tdoor sulfur ra
tios.The es
tima
tes ranged from 0.33
to 0.77 wi
th a medianvalue of 0.53. Ou
tdoor air par
ticles were less impor
tan
tfor personal exposures
than for indoor concen
tra
tions, wi
th
the median ou
tdoor con
tribu
tion
to personal exposurejus
t 49%. We regressed
the ou
tdoor con
tribu
tions
to personalexposures on measured ou
tdoor PM
2.5 a
t the cen
tralsi
te. The regressions had
R2 values ranging from 0.19
to0.88 (median = 0.73). These values provide an indica
tionof
the ex
ten
t of misclassifica
tion error in epidemiologicales
tima
tes of
the effec
t of ou
tdoor par
ticles on heal
th.