Damp housing, mould growth, and symptomatic health state

Abstract

To examine the relation between damp and mould growth and symptomatic ill health.
Cross-sectional study of random sample of households containing children; separate and independent assessments of housing conditions (by surveyor) and health (structured interview by trained researcher).
Subjects' homes (in selected areas of public housing in Glasgow, Edinburgh, and London).
Adult respondents (94% women) and 1169 children living in 597 households.
Specific health symptoms and general evaluation of health among respondents and children over two weeks before interview; and score on general health questionnaire (only respondents).
Damp was found in 184 (30.8%) dwellings and actual mould growth in 274 (45.9%). Adult respondents living in damp and mouldy dwellings were likely to report more symptoms overall, including nausea and vomiting, blocked nose, breathlessness, backache, fainting, and bad nerves, than respondents in dry dwellings. Children living in damp and mouldy dwellings had a greater prevalence of respiratory symptoms (wheeze, sore throat, runny nose) and headaches and fever compared with those living in dry dwellings. The mean number of symptoms was higher in damp and mouldy houses and positively associated with increasing severity of dampness and mould (dose response relation). All these differences persisted after controlling for possible confounding factors such as household income, cigarette smoking, unemployment, and overcrowding. Other possible sources of bias that might invalidate the assumption of a causal link between housing conditions and ill health--namely, investigator bias, respondent bias, and selection bias--were also considered and ruled out.
Damp and mouldy living conditions have an adverse effect on symptomatic health, particularly among children.

Full text

(500
IU)
compared
with
300
[ig
used
in
the
Canadian
work.3
Widespread
administration
of
anti-D
immuno-
globulin
antenatally
in
this
regimen
would
not
be
possible
at
present
because
of
limited
supply
from
a
decreasing
pool
of
immunised
donors.
Immuno-
globulin
produced
by
genetic
engineering,
however,
may
be
available
soon,
and
trials
are
planned
to
study
the
effectiveness
of
even
lower
doses.
When
studying
a
treatment
regimen
for
any
side
effects
it
is
important
to
avoid
the
bias
created
by
considering
only
untoward
consequences.
Unexpected
benefits
are
also
possible,
and
we
paid
particular
attention
to
any
effects
anti-D
immunoglobulin
may
have
had
on
the
incidence
of
hypertensive
disease
such
as
pre-eclampsia.
Some
evidence
suggests
that
previous
blood
transfusions
may
reduce
the
inci-
dence,5
and
possibly
some
blood
products
also
do
so.
The
data
collected,
however,
though
not
contradicting
this
hypothesis,
showed
no
significant
difference.
1
Anonymous.
McMaster
Conference
on
Prevention
of
Rh
Immunisation.
Vox
Sang
1979;36:50-64.
2
Bowman
JM.
Controversies
in
Rh
prophylaxis.
Who
needs
Rh
immune
globulin
and
when
should
it
be
given?
AmJ
Obstet
Gynecol
1985;151:289-94.
3
Bowman
JM,
Chown
B,
Lewis
Mi,
Poll(ck
JM.
Rh
isoimmunization
during
pregnancy:
antenatal
prophylaxis.
Can
Med
,AssocJ
1978;118:623-7.
4
Tovey
LAD.
Haemolvtic
disease
of
the
newborn-the
changing
scene.
BrJ7
Obstet
Gvnaecol
1986;93:960-6.
5
Blaichman
M,
Zipursky
A,
Bartsch
FR,
Hermann
M,
Eklund
J,
Nevanilinna
HR.
Rh
immunization
during
pregnancy.
Vox
Sang
1979;36:50-64.
6
Tovey
LAD,
Townley
A,
Stevenson
BJ,
Taverner
J.
The
Yorkshire
antenatal
anti-D
immunoglobulin
trial
in
primigravidae.
Lancet
1983;ii:244-6.
7
Nusbacher
J,
Bove
JR.
Rh
Immunoprophylaxis:
is
antepartum
therapy
desirable?
N
Engl1J
Med
1980;303:935-7.
8
Tovey
GH.
Should
anti-D
immunoglobulin
be
given
antenatally?
Lancet
1980;ii:466-8.
9
Adams
MM,
Marks
JS,
Kaplan
JP.
Cost
implications
of
routine
antenatal
administration
of
Rh
immune
globulin.
Am
7
Obstet
Gvnecol
1984;149:
633-8.
10
Hensleigh
PA.
Preventing
rhesus
isoimmunization.
Antepartum
Rh
immune
globulin
prophylaxis
versus
a
sensitive
test
for
risk
identification.
Am
J
Obstet
Gynecol
1983;146:749-55.
11
Torrance
GW,
Zipursky
A.
Cost-effectiveness
of
antepartum
prevention
of
Rh
immunisation.
Clin
Perinatol
1984;11:267-81.
12
Hensleigh
PA.
Controversies
in
Rh
prophylaxis.
Am
J
Obstet
Gynecol
1985;153:
112.
13
Bowman
JM.
Controversies
in
Rh
prophylaxis.
AmJ
Obstet
Gynecol
1985;153:
112-3.
14
Tabsch
KMA,
Lebherz
TB,
Crandall
BF.
Risks
of
prophylactic
anti-D
immunoglobulin
after
second-trimester
amniocentesis.
AmJ7
Obstet
Gynecol
1984;149:225-6.
15
Feeney
JG,
Tovey
LAD,
Scott
JS.
Influence
of
previous
blood-transfusion
on
incidence
of
pre-eclampsia.
Lancet
1977;i:874-5.
16
Perrault
RA,
Hogman
CF.
Low
concentration
red
cell
antibodies.
III.
Cold
IgG
anti
D
in
pregnancy,
incidence
and
significance.
Acta
Universitates
Upsaliensis
1972;120:5.
(Accepted
10
April
1989)
Damp
housing,
mould
growth,
and
symptomatic
health
state
Stephen
D
Platt,
Claudia
J
Martin,
Sonja
M
Hunt,
Chris
W
Lewis
Medical
Research
Council
Unit
for
Epidemiological
Studies
in
Psychiatry,
Royal
Edinburgh
Hospital,
Edinburgh
EH10
5HF
Stephen
D
Platt,
PHD,
research
sociologist
Research
Unit
in
Health
and
Behavioural
Change,
University
of
Edinburgh,
Edinburgh
EHI
2QZ
Claudia
J
Martin,
PHD,
research
fellow
Sonja
M
Hunt,
PHD,
senior
research
fellow
Division
of
Applied
Microbiology,
Department
of
Bioscience
and
Biotechnology,
University
of
Strathclyde,
Glasgow
Gl
1XQ
Chris
W
Lewis,
PHD,
research
fellow
Correspondence
to:
Dr
Platt.
Br
MedJ7
1989;298:1673-8
Abstract
Objective-To
examine
the
relation
between
damp
and
mould
growth
and
symptomatic
ill
health.
Design-Cross-sectional
study
of
random
sample
of
households
containing
children;
separate
and
independent
assessments
of
housing
conditions
(by
surveyor)
and
health
(structured
interview
by
trained
researcher).
Setting-Subjects'
homes
(in
selected
areas
of
public
housing
in
Glasgow,
Edinburgh,
and
London).
Subjects-Adult
respondents
(94%
women)
and
1169
children
living
in
597
households.
End
points-Specific
health
symptoms
and
general
evaluation
of
health
among
respondents
and
children
over
two
weeks
before
interview;
and
score
on
general
health
questionnaire
(only
respondents).
Measurements
and
main
results-Damp
was
found
in
184
(30.8%)
dwellings
and
actual
mould
growth
in
274
(45.9%).
Adult
respondents
living
in
damp
and
mouldy
dwellings
were
likely
to
report
more symptoms
overall,
including
nausea
and
vomiting,
blocked
nose,
breathlessness,
backache,
fainting,
and
bad
nerves,
than
respondents
in
dry
dwellings.
Children
living
in
damp
and
mouldy
dwellings
had
a
greater
prevalence
of
respiratory
symptoms
(wheeze,
sore
throat,
runny
nose)
and
headaches
and
fever
compared
with
those
living
in
dry
dwellings.
The
mean
number
of
symptoms
was
higher
in
damp
and
mouldy
houses
and
positively
associated
with
increasing
severity
of
dampness
and
mould
(dose
response
relation).
All
these
differences
persisted
after
controlling
for
possible
confounding
factors
such
as
household
income,
cigarette
smoking,
unemployment,
and
overcrowding.
Other
possible
sources
of
bias
that
might
invalidate
the
assumption
of
a
causal
link
between
housing
con-
ditions
and
ill
health-namely,
investigator
bias,
respondent
bias,
and
selection
bias-were
also
con-
sidered
and
ruled
out.
Conclusion-Damp
and
mouldy
living
conditions
have
an
adverse
effect
on
symptomatic
health,
particularly
among
children.
Introduction
Showing
a
direct
relation
between
damp
housing
and
ill
health
is
by
no
means
straightforward.
Firstly,
those
living
in
the
worst
housing
conditions
are
likely
to
be
experiencing
other
forms
of
adversity,
such
as
low
income
and
unemployment.
Secondly,
personal
behaviour
may
also
play
a
part
in
the
causation
of
ill
health.
An
equally
important
methodological
concern
is
the
process
of
the
data
collection
itself.
If
informa-
tion
about
health
and
housing
conditions
is
elicited
in
the
same
interview
respondents
may
exaggerate
the
prevalence
of
problems,
leading
to
a
spurious
associa-
tion
between
the
two
phenomena.
Moreover,
the
researchers
themselves
may
influence
reporting.
In
1986
we
carried
out
a
preliminary
study
in
Edin-
burgh,
which
attempted
to
overcome
these
methodo-
logical
difficulties
by
using
a
double
blind
research
design.'
Children
living
in
damp
houses,
particularly
where
there
was
also
mould
growth,
were
reported
to
have
higher
rates
of
respiratory
and
gastrointestinal
symptoms,
aches
and
pains,
and
fever
than
children
in
dry
dwellings.
These
differences
could
not
be
attributed
to
smoking
or
differences
between
damp
and
dry
households
regarding
unemployment,
income,
overcrowding,
or
duration
of
tenancy.
The
numbers
of
households
that
included
a
child
was
not
large
enough
(n=
101),
however,
to
permit
a
full
analysis
of
the
role
of
other
possible
confounding
variables.
Accordingly,
we
carried
out
a
larger
scale,
more
detailed
investiga-
tion.
Subjects
and methods
The
study
was
conducted
in
three
major
cities:
Edinburgh,
Glasgow,
and
London.
Within
each
city
discrete
geographical
areas
of
public
housing
were
BMJ
VOLUME
298
24
JUNE
1989
1673

identified
in
which
(a)
families
with
young
children
predominated;
(b)
the
prevalence of
damp
housing
was
thought
to
be
in
the
range
of
25-50%
of
total
dwellings;
(c)
socioeconomic
state
was
likely
to
be
fairly
homo-
geneous;
and
(d)
types
of
housing
and
structures
of
buildings,
including
any
renovations,
could
be
clearly
specified.
Two
sites
were
chosen
in
Edinburgh,
two
in
Glasgow,
and
one
in
London.
Tenants'
groups
were
contacted
and
their
cooperation
elicited.
Lists
of
addresses
at
the
chosen
sites
were
obtained
from
the
relevant
housing
departments.
The
intention
was
to
achieve
a
sample
of
500
eligible
households
in
Edinburgh
and
in
Glasgow
and
200
in
London.
A
random
sample
of
addresses
was
drawn
according
to
the
total
number
of
dwellings
in
the
area.
Only
those
households
with
at
least
one
child
aged
under
16
were
eligible
for
inclusion
in
the
study.
As
official
statistics
on
the
exact
location
of
families
with
young
children
were
not
available
the
sample
was
identified
in
two
ways:
(a)
at
the
time
of
the
main
health
interview
(see
below)
the
interviewers
identified
suit-
able
families
by
contacting
each
dwelling
on
the
list;
and
(b)
in
two
of
the
sites
members
of
the
tenants'
association
identified
addresses
on
the
list
containing
families
who
met
the
study
criteria.
Two
surveyors
carried
out
an
assessment
of
damp-
ness
(severity
and
type)
and
mould
(severity
and
location)
and
details
of
the
structure
of
the
dwelling.
Using
an
air
sampler
(Surface
Air
Systems)
they
extracted
air
samples
from
rooms
and,
where
visible
mould
growth
was
present,
a
sample
from
each
affected
room
was
collected.
A
microbiologist
esti-
mated
spore
counts
from
the
air
samples
and
identified
the
fungi
from
air
and
walls
when
possible.
We
devised
and
pretested
two
survey
forms.
The
form
for
the
house
conditions
survey
contained
items
on
type
of
building,
location,
number
of
rooms,
dampness,
mould,
ventilation, insulation,
and
reno-
vations.
The
health
survey
was
a
revised
version
of
that
used
by
Martin
et
al.'
In
the
course
of
a
structured
interview
the
respondent
(whenever
possible
the
female
householder)
answered
detailed
questions
about
her
own
and
her
children's
health
during
the
past
two
weeks;
smoking
by
all
adults
and
children;
type
of
heating,
washing,
and
drying
facilities;
presence
of
pets;
economic
activity
and
occupation
of
all
adults
in
the
household;
household
income;
and
housing
conditions
and
facilities.
The
study
was
carried
out
during
February-April
1988.
Once
the
health
interview
had
been
completed
the
surveyors
were
instructed
to
visit
the
dwelling.
The
petri
dishes
containing
air
and
wall
mould
samples
were
taken
each
day
to
the
University
of
Strathclyde,
where
they
were
refrigerated
and
cultured.
Air
spore
counts
were
calculated
and
fungi
identified
when
possible.
The
surveyors
and
the
microbiologist
were
blind
to
each
other's
findings
and
also
to
the
findings
of
the
health
survey
team.
We
used
four
categorical
independent
variables
re-
lating
to
housing
conditions.
Households
that
received
a
house
conditions
survey
were
classified
into
three
groups:
those
where
there
was
no
objective
evidence
of
dampness
or
mould
growth
(dry),
those
with
only
damp,
and
those
with
mould
(whether
or
not
dampness
was
also
present).
The
overall
dampness
in
the
house-
hold
was
calculated
by
averaging
the
score
for
each
bedroom,
sitting
room,
and
kitchen
on
a
four
point
scale
of
severity
(0=none;
3=severe).
Households
in
which
the
average
dampness
score
exceeded
zero
(no
dampness
whatsoever)
were
divided
into
three
approximately
equal
sized
groups
labelled
mild
(score
ranging
between
0-01
and
0
52),
moderate
(0
53
to
1
05),
and
severe
(u
1-06).
A
similar
procedure
was
adopted
to
divide
households
into
four
groups
differ-
ing
in
average
severity
of
mould
(none,
mild
(0
01
to
0-45),
moderate
(0
46
to
0
77),
and
severe
(¢0
78)).
The
spore
concentration
per
m'
air
was
measured
in
the
kitchen,
living
room,
and
bedrooms
of
households
in
Edinburgh
and
Glasgow
visited
by
the
surveyors.
On
the
basis
of
preliminary
work
in
Edinburgh
(B
Flanagan
and
C
A
Hunter,
unpublished
data)
and
elsewhere2-4
we
devised
a
five
point
scale
(coded
1
to
5):
low
(
100
viable
spores/m3
air),
medium
(101-300),
high
(301-1000),
very
high
(1001-5000),
and
extremely
high
(>5000).
The
household
spore
concentration
was
the
mean
score
on
the
scale
per
available
room.
A
new
variable
was
created
by
dividing
this
mean
score
into
three
groups:
low
(scoring
1),
medium
(1
01
to
2
00),
and
high
(>2).
To
ensure
that
the
relation
between
housing
con-
ditions
and
ill
health
was
not
invalidated
by
covariation
with
other
variables
several
possible
confounding
factors
were
also
examined,
particularly
cigarette
smoking
in
the
household
(no/yes),
respondents'
cigar-
ette
smoking
(no/yes),
net
household
income
(above
median
(£80)/below
median),
overcrowding
(less
than/
more
than
1-5
people
per
room),
employment
in
the
household
(somebody
employed/nobody
employed),
and
employment
state
of
the
respondent
(employed/
unemployed,
no
paid
employment).
The
respondent
was
asked
to
report
on
the
presence
of
16
specific
symptoms
seen
in
the
past
two
weeks
in
any
child
(aged
0-15)
living
in
the
household.
We
devised
two
summary
symptom
scores
relating
to
children:
the
unadjusted
score
being
the
total
number
of
symptoms
among
all
children
in
the
household
and
the
adjusted
score
being
the
total
of
symptoms
divided
by
the
number
of
children-that
is,
the
mean
number
of
symptoms
per
child.
Another
summary
dependent
variable
for
children
was
the
mean
score
on
health
evaluation
derived
from
the
respondent's
general
evaluation
of
each
child
on
a
scale
of
1
(excellent)
to
5
(very
poor).
The
respondent
was
also
asked
to
report
whether
she
had
suffered
from
any
of
17
specific
symp-
toms
over
the
past
fortnight.
A
summary symptom
score
wvas
merely
the
sum
of
individual
symptoms.
In
addition,
the
respondent
was
asked
to rate
her
general
health
on
the
same
five
point
scale
used
for
children
and
to
complete
the
30
item
general
health
question-
naire5
(range
0-30),
here
used
as
a
general
indicator
of
psychological
distress.
Finally,
we
inquired
about
medical
treatment
for
symptoms
and
the
presence
of
a
recurrent
or
longstanding
illness
among
both
re-
spondents
and
children.
Univariate
analyses
of
the
relation
between
each
independent
variable
and
dependent
variables
were
carried
out
with
x)
tests
(categorical
variables)
or
one
way
analysis
of
variance
(metric
variables).
Subse-
quently,
multivariate
analyses
were
performed
to
examine
the
association
between
housing
conditions
and
ill
health
after
controlling
for
possible
confounding
factors.
When
the
response
variable
was
binary/
categorical
we
used
logistic
linear
regressiQn
analysis6;
for
metric
response
variables
we
used
analysis
of
covariance.7
The
extent
of
any
dose-response
relation
between
severity
of
dampness,
mould
growth,
and
air
spore
concentration
and
health
was
assessed
by
means
of
tau
c
(categorical
variables)
and
the
Pearson
correlation
coefficient
(metric
variables).
Identical
results
were
obtained
with
respect
to
metric
variables
transformed
to
base
10
logarithms.
Only
original
values
are
reported
below.
On
the
basis
of
previous
work
we
expected
to
find
a
distinct
effect
of
adverse
housing
conditions
on
respira-
tory
and
gastrointestinal
symptoms
in
children
and
on
emotional
distress
in
adults.
Evidence
of
a
dose
response
relation
was
considered
to
be
particularly
relevant
in
assessing
the
likelihood
of
a
causal
impact
of
dampness
and
mould
on
symptomatic
health.
For
the
purposes
of
this
report
the
results
from
BMJ
VOLUME
298
24
JUNE
1989
1674

JEdinburgh,
Glasgow,
and
London
have
been
com-
bined.
(Although
the
prevalence
of
damp
and
mould
varied
in
each
city,
there
were
no
pronounced
differ-
ences
in
the
association
between
housing
conditions
and
symptomatic
health
state
between
cities.)
Results
RATES
OF
RESPONSE
Of
1220
households
with
children
eligible
for
inclu-
sion
in
the
study,
a
health
interview
was
secured
in
891
(73
0%);
156
(12-8%)
respondents
refused
to
be
inter-
viewed,
and
173
(14
2%)
could not
be
contacted.
Surveyors
completed
their
investigations
of-
housing
conditions
in
597
households,
constituting
48
9%
of
eligible
households
and
67-0%
of
those
who
had
the
health
interview.
A
comparison
between
surveyed
(n=597)
and
non-surveyed
(n=294)
households
showed
no
differences
in
sociodemographic
charac-
teristics,
such
as
gender,
marital
state,
household
size
(including
number
of
children),
social
class,
and
overcrowding,
or
regarding
disposable
income,
cigar-
ette
smoking,
length
of
time
at
current
address,
presence
of
pets,
or
self-reported
damp
or
mould.
The
only
significant
difference
concerned
employment:
131
(22%)
respondents
in
surveyed
households
were
employed
compared
with
100
(34%)
respondents
in
non-surveyed
households
(x2=
12-54,
df=
1,
p<0Q001);
corresponding
figures
for
any
adult
in
employment
were
257
(43%)
and
156
(53%),
respectively
(x2=7-55,
df=1,
p<001).
All
subsequent
analyses
were
based
on
the
597
households,
containing
1169
children,
that
received
both
a
housing
survey
and
a
health
interview.
COMPARISON
OF
THREE
HOUSING
CONDITIONS
GROUPS
Out
of
the
597
households,
only
184
(30
8%)
were
free
from
damp
or
mould
(dry).
In
139
(23-3%)
households
surveyors
found
evidence
of
damp
and
in
274
(45
9%,
of
which
all
but
nine
were
also
damp)
actual
mould
growth
was
visible.
The
three
housing
conditions
groups
(dry,
only
damp,
mouldy)
were
compared
for
descriptive
purposes
on
a
number
of
background
(sociodemographic
and
other)
variables.
(It
was,
of
course,
recognised
that
a
variable
could
act
as
a
confounder
even
if
it
did
not
differentiate
significantly
between
groups.)
Only
one
significant
difference
emerged:
respondents
living
in
dry
house-
holds
had
been
living
an
average
of
5
9
(SD
4
9)
years
at
the
address
compared
with
4-8
(4-1)
years
among
respondents
in
damp
houses
and
6
4
(5
5)
years
among
TABLE
I-Respondent's
health
during
past
two
weeks
by
housing
conditions.
Figures
are
numbers
(percentages)
unless
stated
otherwise
Housing
conditions
Significance
No
damp
Degrees
or
mould
Damp
only
Mould
of
p
Symptom
(n=
184)
(n=
139)
(n=274)
y
freedom
Value
Tiredness
76(41
3)
69(50-0)
141
(51-5)
4.-4
2
0
089
High
bloodpressure
9(4-9)
7(5-1)
22(8-0)
2-33
2
0-312
Persistentcough
30(16-3)
27(19-4)
64(23-4)
3-47
2
0-177
Bad
nerves
35
(19-0)
31
(22-3)
80
(29-2)
6-62
2
0-036
Wheezing
19(10-3)
17(12-2)
37(13
6)
1-07
2
0-587
Aching
joints
28
(15
2)
23
(16-5)
65
(23-7)
6-05
2
0-049
Skinprohlems
26(14-1)
23(16-5)
43(15
7)
0-39
2
0-825
Persistent
headaches
49
(26-6)
43
(30-9)
75
(27
4)
0-82
2
0
664
Nausea-vomiting
7(3-8)
9(6-5)
27
(9-9)
6
17
2
0-046
Backache
41(22-3)
48(34-5)
81(29
6)
6-13
2
0
047
Blocked
nose
25
(13
6)
18(12-9)
5821
-2)
6-53
2
0-038
Palpitations
8
(4-3)
9
(6-5)
22
(8-0)
2-44
2
0-295
Fainting
spells
3(1-6)
12
(8-6)
17(6-2)
8-37
2
0-015
Diarrhoea
5
(2-7)
9(6-5)
19(6-9)
4-06
2
0-131
Constipation
11
(6-0)
8
(5-8)
33
(12-0)
7-08
2
0-029
Breathlessness
19(10-3)
24(17-3)
51(18-6)
6-01
2
0-049
Feeling
depressed
51
(27-7)
47
(33-8)
104
(38-0)
5-15
2
0-076
Anysymptom
144(78-3)
113(81-3)
217(79-2)
0-46
2
0-795
Mean
(SD)
No
of
symptoms
2-40
(2-37)
3-05
(3-01)
3-43
(3-25)
F=6-67
2,594
0-001
Mean
(SD)
health
evaluation
score
2-41
(0-93)
2-49
(0-99)
2-66
(0-97)
F=4-09
2,594
0-017
Mean
(SD)
general
health
questiontsaire
score
5-74
(7-
12)
6-87
(7-78)
7-20
(8-35)
F=
1-92
2,583
0-148
respondents
in
mouldy
houses
(F=4-35,
df=2,584,
p<002);
only
the
difference
between
damp
and
mouldy
houses
was
significant
(Scheffe
test,
p<0
05).
Housing
groups
did
not
differ
in
number
of
children
(mean
(SD)
2-0
(1
0)),
total
number
of
household
members
(3-8
(1
-2)),
respondent's
gender
(559
(93
6%)
women),
respondent's
marital
state
(384
(64
3%)
married),
net
household
income
(293
(49
0%)
under
£80
per
week),
respondent's
smoking
(415,
(69
5%)),
any
smoker
in
household
(476
(79-7%)),
respondent
employed
(136
(22-8%)),
any
household
member
employed
(259
(43
4%)),
overcrowding
(109
(18
3%)),
presence
of
pets
(269
(45-
1%)),
tenure
of
last
house
(465
(77
9%)
council
dwelling),
reasons
for
moving
from
last
dwelling
(90
(15
0%)
because
of
dampness;
247
(41
4%)
because
of
other
problems
with
the
house;
26
(4-3%)
for
health
reasons),
and
use
of
Calor
gas
heating
(81
(13-5%)).
Respondents
in
mouldy
house-
holds,
however,
reported
more
problems
apart
from
the
damp
(especially
noise,
poor
repair,
and
cold)
than
respondents
in
damp
or
dry
households.
(Mean
(SD)
problems
2
7
(1
5),
2
5
(1-6),
and
2-2
(1
6),
respec-
tively;
F=5-0,
df=2,594,
p<0
01).
In
particular,
the
prevalence
of
cold
as a
problem
was
reported
in
222
(81%),
100
(72%),
and
114
(62%)
households,
respectively;
2
=20
4,
df=2,
p<0
001).
HOUSING
CONDITIONS
AND
RESPONDENT'S
HEALTH
Table
I
shows
the.
relation
between
prevalence
of
symptoms
in
the
respondent
and
housing
conditions.
Significant
differences
between
groups
were
found
regarding
bad
nerves,
aching
joints,
nausea
and
vomiting,
backache,
blocked
nose,
fainting
spells,
constipation,
and
breathlessness.
The
lowest
propor-
tion
reporting
symptoms
was
found
in
dry
households;
with
only
one
exception
(fainting
spells)
the
highest
proportion
was
found
in
mouldy
households.
Al-
though
housing
conditions
were
unrelated
to
the
presence
of
any
particular
symptom,
there
was
a
significant
variation
in
the
total
number
of
symptoms
and
in
the
respondent's
evaluation
of
her
health.
In
particular,
those
living
in
mouldy
houses
scored
significantly
higher
than
those
living
in
dry
conditions
(Scheffe
test,
p<0
05).
The
general
health
question-
naire
score
was
not
related
to
housing
conditions
(table
I).
Preliminary
univariate
analyses
had
shown
that
only
two
of
the
possible
confounding
variables
(re-
spondent's
economic
position
and
cigarette
smoking)
were
significantly
associated
with
the
presence
or
absence
of
individual
symptoms.
We
therefore
under-
took
a
series
of
logistic
regression
analyses
in
which
the
dependent
variables
were
the
eight
symptoms
previously
shown
to
be
significantly
associated
with
housing
conditions.
After
controlling
for
the
respond-
ent's
economic
position
and
cigarette
smoking
these
differences
remained
significant
for
all
eight
dependent
symptom
variables
(problem
free
households
always
having
the
lowest
proportion
of
respondents
positive
for
symptoms).
The
relation
between
housing
conditions
on
the
one
hand
and
the
total
number
of
symptoms,
health
evaluation
score,
and
general
health
questionnaire
score
on
the
other
was
further
examined
by
means
of
analyses
of
covariance.
After
we
controlled
for
length
of
time
at
address,
other
housing
problems
(or
cold
alone),
respondent's
economic
position,
respondent's
cigarette
smoking,
and
household
income
housing
conditions
remained
significantly
associated
with
the
total
number
of
symptoms
(6
ranging
between
0-10
and
0
14,
p<0
05
to
<0
005),
with
those
living
in
mouldy
households
reporting
most
and
those
in
dry
households
fewest
symptoms.
Housing
conditions
were
not
significantly
associated
with
health
evaluation
score
after
we
controlled
for
other
possible
con-
BMJ
VOLUME
298
24
JUNE
1989
1675
1675

founding
variables,
and
the
relation
with
the
general
health
questionnaire
score
remained
non-significant.
We
examined
the
dose-response
relation
between
the
respondents'
symptoms
and
increasing
severity
of
dampness,
mould
growth,
and
air
spore
concentration.
Table
II
summarises
the
findings
of
these
analyses.
TABLE
iI-Respondent's
health
during
past
two
weeks.
Dose-response
relation
with
damp,
mould,
and
air
spore
count.
Figures
are
tau
c
values
(p
values)
unless stated
otherwise
Dampness
Mould
growth
Air
spore
count
Symptom
(Max
n=
597)
(Max
n=589)
(Max
n=485)
Tiredness
0-09
(0-028)
0-06
(0-076)
-0-02
(0-341)
High
blood
pressure
0-04
(0-024)
0-04
(0-027)
0-05
(0-017)
Persistent
cough
0-09
(0-010)
0-04
(0-
110)
0-06
(0-062)
Bad
nerves
0-07
(0-036)
0-09
(0-008)
0-08
(0-031)
Wheezing
0-05
(0-047)
0-03
(0-
125)
0-01
(0-413)
Aching
joints
0-05
(0-080)
0-07
(0-022)
0-06
(0-083)
Skin
problems
0-03
(0-209)
0-00
(0-474)
0-06
(0-063)
Persistent
headaches
0-04
(0-
150)
-0-02
(0-279)
-0-11
(0
006)
Nausea-vomiting
0-04
(0-044)
0-05
(0-015)
0-02
(0
230)
Backache
0-04(0-167)
0-02(0-332)
0-11
(0-009)
Blocked
nose
0-11
(0-001)
0-08
(0-005)
0-00(0-451)
Palpitations
0-03
(0
096)
0-03
(0-051)
0
08
(0-001)
Fainting
spells
0-05
(0-013)
0-01
(0-381)
-0-01
(0-289)
Diarrhoea
0-02
(0-146)
0-02
(0-
109)
-0-01
(0-413)
Constipation
0-02
(0-271)
0-04
(0-054)
0-01
(0-414)
Breathlessness
0-09
(0-003)
0-05
(0-057)
0-08
(0-019)
Feeling
depressed
0-06
(0-081)
0-08
(0-026)
0-06
(0-
107)
Anv
symptom
0-02
(0-319)
-0-02
(0-299)
0-00
(0-482)
No
of
symptoms
r=0-
14
(0-001)
r=0-09
(0-014)
r=0-08
(0-039)
Health
evaluation
score
r=0-07
(0-047)
r=0-
10
(0-008)
r=0-05
(0-
115)
General
health
questionnaire
r=0-06
(0-082)
r=0-06
(0-086)
r=0-01
(0-414)
There
was
a
significant
tendency
for
increasing
severity
of
dampness
to
be
associated
with
a
greater
prevalence
of
the
following
symptoms:
tiredness,
high
blood
pressure,
persistent
cough,
bad
nerves,
wheezing,
nausea
and
vomiting,
blocked
nose,
fainting
spells,
and
breathlessness.
The
greater
the
extent
of
mould
growth
the
higher
the
proportion
of
respondents
reporting
high
blood
pressure,
bad
nerves,
aching
joints,
nausea
and
vomiting,
blocked
nose,
and
feeling
depressed.
Finally,
the
concentration
of
the
air
spores
was
positively
associated
with
high
blood
pressure,
bad
nerves,
backache,
palpitations,
and
breathlessness
and
negatively
associated
with
persistent
headaches.
Overall,
the
total
number
of
symptoms
tended
to
increase
with
higher
degrees
of
dampness
and
mould
and
air
spore
concentration,
while
the
health
evalua-
tion
score
was
related
only
to
severity
of
dampness
and
mould
growth.
No
dose-response
effect
on
the
general
health
questionnaire
score
was
evident.
Respondents
living
in
the
three
different
housing
conditions
were
compared
regarding
action
taken
TABLE
III-Children's
health
during
past
two
weeks
by
housing
conditions.
Figures
are
number
(percentages)
unless
stated
otherwise
Housing
conditions
Significance
No
damp
Damp
Degrees
or
mould
only
Mould
of
p
Symptom*
(n=
184)
(n=
139)
(n=274)
X
freedom
Value
Bodily
aches-pains
23
(12-5)
30(21-6)
43
(15-7)
4-90
2
0-086
Diarrhoea
34
(18
5)
30
(21-6)
50
(18-2)
0-73
2
0-694
Wheezing
30(16
3)
26(18-7)
74(27-0)
8-41
2
0-015
Vomiting
22(12-0)
25(18-0)
52(19-0)
4-18
2
0-124
Sorethroat
56(30-4)
34(24-5)
116(42-3)
14-99
2
<0-001
Irritability
23(12-5)
28(20-1)
56(20-4)
5
-32
2
0-070
Tiredness
25(13-6)
28(20-1)
48(17-5)
2-55
2
0-279
Persistentheadaches
23(12-5)
19(13-7)
58(21-2)
7-16
2
0-028
Earache
27
(14-7)
15
(10-8)
47(17-2)
2-95
2
0-228
Fever-high
temperature
21(11-4)
25
(18-0)
67
(24-5)
12-30
2
0-002
Feelingdepressed-unhappy
20(10-9)
25
(18-0)
42
(15-3)
3-45
2
0-178
Tempertantrums
37(20-1)
37(26-6)
74(27-0)
3-13
2
0-209
Bedwetting
41(22-3)
29
(20-9)
64
(23-4)
0-33
2
0-846
Poor
appetite
31
(16-8)
37
(26-6)
68
(24-8)
5-49
2
0-064
Persistent
cough
57
(31-0)
52
(37-4)
117
(42-7)
6-45
2
0-040
Runnynose
72(39-1)
56(40-3)
139(50-7)
7-43
2
0-024
Anysymptom
147(79-9)
119(85-6)
248(90-5)
10-41
2
0-006
Mean(SD)Noofsymptoms
3-73(3-95)
4-39(4-63)
5-44(5-19)
F=7-56
2,594
<0-001
Mean
(SD)
No
of
symptoms
per
child
2-04(1-98)
2-46
(2-36)
2-86
(2-43)
F=7-23
2,594
<0-001
Mean
(SD)
health
evaluation
score
2-24
(0-89)
2-30
(0-91)
2-41
(0-94)
F=
1-98
2,592
0-140
*Symptom
present
in
any
child
living
in
household.
during
the
past
two
weeks
to
deal
with
symptoms
and
presence
of
recurrent
and
long-standing
illness.
No
significant
differences
were
found.
HOUSING
CONDITIONS
AND
CHILDREN
S
HEALTH
Table
III
shows
the
prevalence
of
symptoms
among
children
in
the
household
by
housing
conditions.
Significant
differences
were
found
regarding
wheezing,
sore
throat,
persistent
headache,
fever
and
high
tem-
perature,
persistent
cough,
and
runny
nose.
The
highest
proportion
reporting
these
symptoms
was
always
found
in
mouldy
households;
with
only
one
exception
(sore
throat)
the
lowest
proportion
with
symptoms
was
found
in
the
dry
households.
Not
only
was
there
a
significant
difference
in
the
proportion
with
any
symptom
(147
(79
9%)
in
dry
households,
119
(85
6%)
in
damp
houses,
248
(90
5%)
in
mouldy
houses)
but
the
mean
number
of
symptoms
(overall
and
per
child)
also
differed
significantly
and
in
the
same
direction.
The
mean
child
health
evaluation
score
was
not
significantly
different
between
groups
(table
III).
In
our
preliminary
univariate
analyses
we
had
noted
that
three
of
the
possible
confounding
variables
(over-
crowding,
any
cigarette
smoker,
nobody
employed)
were
significantly
associated
with
presence
or
absence
of
individual
symptoms.
Another
set
of
logistic
regres-
sion
analyses
was
therefore
undertaken
in
which
the
dependent
variables
were
the
six
symptoms
previously
shown
to
be
significantly
associated
with
housing
conditions.
After
controlling
for
these
three
con-
founding
variables
differences
remained
significant
for
wheezing,
sore
throat,
persistent
headache,
fever
and
high
temperature,
runny
nose,
and
for
any
symptoYn.
Only
the
main
effect
of
housing
conditions
on
cough
was
no
longer
significant.
The
relation
between
housing
conditions
on
the
one
hand
and
mean
number
of
symptoms
and
mean
health
evaluation
score
on
the
other
was
further
examined
in
a
series
of
analyses
of
covariance.
As
before,
we
took
into
account
differences
in
the
length
of
time
at
address
and
other
housing
problems
(or
cold
alone).
We
also
added
a
control
for
the
number
of
children
in
the
household
and
the
adult's
general
health
questionnaire
score
(included
because
although
it
did
not
differ
signifi-
cantly
with
housing
conditions,
it
was
correlated
highly
with
both
the
mean
number
of
symptoms
in
children
(r=0
30,
p<0
001)
and
mean
child
evaluation
score
(r=0-35,
p<0-001)).
Finally,
we
partialled
out
the
effects
of
cigarette
smoking
in
the
household,
unemployment,
low
income,
and
overcrowding.
There
was
still
a
significant
effect
of
housing
conditions
on
the
mean
number
of
symptoms
(6
ranging
between
0
10
and
0
13,
p<002
to
<0
005).
Children
living
in
mouldy
households
were
reported
to
have
the
highest
number
of
symptoms
and
those
living
in
dry
house-
holds
the
fewest.
Mean
child
evaluation
score
remained
unrelated
to
housing
conditions.
Table
IV
shows
the
dose-response
relation
between
children's
symptoms
and
increasing
severity
of
damp-
ness,
mould
growth,
and
air
spore
concentration.
The
more
serious
the
dampness
the
greater
the
prevalence
of
bodily
aches
and
pains,
wheezing,
vomiting,
sore
throat,
irritability,
tiredness,
persistent
headache,
fever
and
high
temperature,
feeling
depressed
and
unhappy,
poor
appetite,
persistent
cough,
and
runny
nose.
Dampness
was
also
associated
overall
with
the
presence
of
any
symptom.
The
more
severe
the
mould
growth
the
greater
the
likelihood
of
wheezing,
sore
throat,
irritability,
persistent
headache,
fever
and
high
temperature,
and
runny
nose.
Mould
growth
was
also
associated
with
the
presence
of
any
symptom.
The
greater
the
air
spore
concentration
the
greater
the
prevalence
of
wheezing,
irritability,
and
fever
and
high
temperature.
BMJ
VOLUME
298
24
JUNE
1989
1676

TABLE
IV
-Children's
health
during
past
two
weeks.
Dose-response
relation
with
damp,
mould,
and
air
spore
count.
Figures
are
tau
c
values
(p
values)
unless
stated
otherwise
Dampness
Mould
growth
Air
spore
count
Symptom
(Max
n=597)
(Max
n=589)
(Max
n=485)
Bodily
aches-pains
0-08
(0-006)
-0-01
(0-383)
-0-01
(0
384)
Diarrhoea
0-02
(0-291)
-0-01(0-386)
0-01(0-361)
Wheezing
0-10
(0-005)
0-09
(0-005)
0-07
(0
044)
Vomiting
0-06
(0-029)
0-04
(0-106)
0
03
(0
238)
Sore
throat
0-09
(0-020)
0-
14
(<0
001)
0-03
(0-264)
Irritability
0-10
(0-004)
0-06
(0
040)
0
07
(0-033)
Tiredness
0-06(0-043)
0-01(0
365)
0-01(0-351)
Persistent
headaches
0-
12
(<0-001)
0-09
(0-002)
0-00
(0-456)
Earache
-0-01
(0-349)
0-03
(0-170)
-0-04
(0-
130)
Fever-high
temperature
0-12
(<0-001)
0-10
(0-002)
0-06
(0-046)
Feeling
depressed-
unhappy
0-08
(0-007)
0-02
(0
237)
-0-02
(0-294)
Temper
tantrums
004
(0-159)
0-06
(0
069)
0-01
(0-399)
Bedwetting
0-02
(0-313)
0-00
(0
460)
-0-01
(0
437)
Poor
appetite
0-08
(0-015)
0
03
(0-200)
0-02
(0-336)
Persistent
cough
0-11
(0-006)
0-06
(0-068)
0-05
(0-
139)
Runny
nose
0-08
(0-033)
0-09
(0-023)
0-06
(0
123)
Any
symptom
0-08
(0-005)
0-07
(0-011)
0-00
(0-492)
MeanNoofsymptoms
r=0-17(0-001)
r=0-14(0-001)
r=0-11(0
010)
Mean
No
of
symptoms
per
child
r=0-13
(0
001)
r=0-
12(0-002)
r=0-05
(0-161)
Mean
health
evaluation
score
r=0
08
(0
025)
r=0-07
(0-044)
r=0-06
(0-107)
Overall,
the
mean
number
of
symptoms
tended
to
increase
with
greater
severity
of
dampness,
mould
growth,
and
air
spore
concentration,
whereas
the
mean
number
of
symptoms
per
child
and
the
mean
child
health
evaluation
score
were
related
only
to
greater
doses
of
dampness
and
mould
growth.
The
mean
number
of
symptoms
per
child
and
the
mean
child
health
evaluation
score
were
unrelated
to
the
extent
of
air
spore
concentration.
The
three
groups
of
housing
conditions
were
com-
pared
regarding
the action
taken
to
deal
with
children's
symptoms
during
the
past
two
weeks
and
presence
of
recurrent
and
longstanding
illness.
Children
in
mouldy
households
were
more
likely
to
have
been
given
medicines
(51-8%)
than
children
in
damp
(43-2%)
or
problem
free
households
(36-4%)
(X2=
1082,
df=2,
p<0
005).
Other
differences
did
not
reach
significance.
Discussion
Before
offering
an
account
of
the
role
of
damp
and
mould
in
the
aetiology
of
symptoms
it
is
necessary
to
consider
four
types
of
bias
that
may
invalidate
the
assumption
of
a
causal
link
between
housing
con-
ditions
and
ill
health-namely,
investigator
bias,
respondent
bias,
selection
bias,
and
omitted
variable
bias.
Investigator
bias
may
be
dismissed
as
housing
conditions
and
the
health
of
household
members
were
independently
assessed
by
two
different
groups
of
researchers,
neither
of
which
included
the
principal
investigators.
In
addition,
questionnaires
were
coded
and
data
prepared
by
workers
who
were
not
familiar
with
the
objectives
of
the
study.
Some
previous
investigations
of
Ye
housing-health
relation,
particularly
those
carried
out
by
tenants'
groups,
have
been
criticised
on
the
grounds
that
people
living
in
damp
and
mouldy
houses
will
be
inclined
to
exaggerate
the
extent
of
their
own
and
their
children's
health
problems.
A
recent
study
suggested
that
the
observed
association
between
mould
and
respiratory
symptoms
may
be
accounted
for
by
parental
awareness
of
mould
in
the
home.'
Our
reliance
on
informants'
reports
about
the
health
of
themselves
and
their
children
was
deliberate.
We
were
unconvinced
about
the
reliability
and
appropriateness
of
diagnostic
data
derived
from
official
records,
especially
those
of
general
practitioners.
We
thought
that
it
was
valid
to
assess
health
state
by
means
of
self
reported
symptoms
while
at
the
same
time
recognising
that
the
likelihood
of
respondent
bias
was
thereby
increased.
This
problem
was
minimised,
however,
by
the
use
of
inde-
pendent,
expert
assessments
of
housing
conditions.
Although
subjective
(self
reported)
and
objective
(expert)
evaluations
of
the
presence
of
damp
and
mould
were
significantly
and
positively
associated
(k=0
26,
p<0O001),
there
was
disagreement
about
damp
and
mould
state
in
183
(30
7%)
of
the
dwellings.
Furthermore,
respondents
could
not
have
been
aware
of
the
air
spore
concentration
in
the
building.
(The
association
between
self
reported
damp
mould
and
spore
count,
although
significant,
was
not
high:
r=014,
p<0001.)
Nevertheless,
symptoms
in
both
children
and
respondents
were
related
to
this
measure.
We
also
included
the
general
health
questionnaire
score
as
a
covariate
when
examining
the
effect
of
housing
conditions
among
children
as
respondents
with
greater
levels
of
psychological
distress
tended
to
report
more
ill
health.
The
mean
number
of
symptoms
remained
significantly
higher
in
damp
and
mouldy
dwellings
than
in
dry
dwellings.
Thus
though
the
overall
number
of
symptoms
may
have
been
higher
than
would
be
obtained
by
an
independent
observer,
there
is
no
reason
to
believe
that
such
a
bias
affected
the
main
findings.
Another
possible
source
of
error
is
that
of
selection
bias.
People
who
already
suffer
from
ill
health
may
tend
to
live
in
damp
or
mouldy
dwellings:
symptoms
may
exist
before,
rather
than
be
a
consequence
of,
living
in
poor
housing
conditions.
This
could
happen,
for
example,
where
the
least
desirable
dwellings
were
allocated
to
those
most
in
need
who, by
virtue
of
low
income,
social
circumstances,
or
medical
history,
were
more
likely
to
report
ill
health.
Although
housing
departments
may
not
always
act
impartially
in
the
selection
of
tenants
to
households,
there
is
no
evidence
to
suggest
that
they
systematically
allocate
families
in
poorer
health
to
damp
and
mouldy
households.
In
this
study
families
in
damp
and
mouldy
dwellings
were
not
more
likely
to
have
come
from
previously
poor
con-
ditions
or
to
have
moved
for
health
reasons
or
to
have
lived
a
shorter
period
of
time
in
the
dwelling
than
families
in
dry
houses.
In
addition,
many
of
the
children
in
all
three
housing
groups
were
born
in
the
household
in
which
they
were
currently
living.
Thus
selection
bias
is
highly
unlikely
to
account
for
the
findings.
Omitted
variable
bias
can
arise
when
variables
that
are
correlated
with
the
major
independent
variable
(in
this
case
housing
conditions)
and
have
a
significant
(possibly
causal)
relation
with
the
dependent
(outcome)
variable
(such
as
symptom
score)
are
excluded
from
the
analysis.
Whereas
several
factors
were
significantly
associated
with
health
state,
only
cold
was
also
asso-
ciated
with
housing
conditions.
Cold
stress
may
have
made
some
contribution
to
the
experience
of
symptoms:
a
damp
house
is
usually
a
cold
house.
Unfortunately,
we
were
unable
to
assess
the
tempera-
ture
of
dwellings.
We
did,
however,
gather
informa-
tion
on
perceived
coldness
of
the
dwelling
and
this
variable
was
included
in
the
covariance
analysis.
In
summary,
adult
respondents
living
in
damp
and
mouldy
dwellings
were
more
likely
to
report
nausea,
vomiting,
constipation,
blocked
nose,
breathlessness,
backache,
aching
joints,
fainting,
and
bad
nerves
than
respondents
living
in
dry
dwellings.
These
differences
remained
after
controlling
for
the
respondent's
economic
position
and
cigarette
smoking.
In
a
more
extensive
covariance
analysis
respondents
living
in
mouldy
dwellings
were
found
to
have
the
highest
number
of
symptoms
even
after
taking
account
of
possible
confounding
factors
such
as
length
of
time
at
address,
other
housing
problems,
household
income,
economic
position,
and
cigarette
smoking.
This
analysis,
however,
showed
that
the
respondent's
sub-
jective
evaluation
of
health
and
psychological
distress
BMJ
VOLUME
298
24
JUNE
1989
1677

were
both
unrelated
to
housing
conditions.
Increasing
doses
of
dampness
and
mould
were
especially
linked
to
nausea,
blocked
nose,
breathlessness,
high
blood
pressure,
and
bad
nerves
and
to
a
greater
number
of
symptoms
and
a
poorer
health
evaluation
score.
For
children,
living in
damp
and
mouldy
dwellings
was
associated
with
a
greater
prevalence
of
wheeze,
sore
throat,
runny
nose,
cough,
headaches,
and
fever
compared
with
those
living
in
dry
dwellings.
With
the
exception
of
cough
these
differences
were
unaffected
by
the
introduction
of
controls
for
smoking
in
the
household,
employment,
and
overcrowding.
Addi-
tional
possible
confounding
variables
were
added
in
an
analysis
of
covariance,
which
still
showed
a
significant
effect
of
housing
conditions
on
the
mean
number
of
symptoms
among
children
in
the
household.
A
dose-
response
relation
was
particularly
noted
with
respect
to
wheeze,
sore
throat,
runny
nose,
irritability,
persistent
headache,
and
fever
and
high
temperature.
Increasing
severity
of
dampness
and
mould
and
any
symptom,
the
mean
number
of
symptoms
(overall
and
per
child),
and
the
mean
child
health
evaluation
score
were
also
associated.
Several
studies
have
suggested
that
some
varieties
of
fungal
spores
are
allergenic
and
give
rise
to
respiratory
conditions.
Burr
et
al
identified
Penicillium
notatum,
Cladosporium
herbarum,
and
Aspergillus
species
in
the
homes
of
asthmatic
patients
and
found
that
the
moulds
gave
positive
skin
test
reactions
for
allergy.9
Fungal
spores
are
also
believed
to
affect
the
respiratory
tract
by
producing
tissue
lesions,
by
forming
saprophytic
colonies
on
mucus
plugs,
and
by
causing
inflammation
and
irritation
of
nasal
and
bronchial
passages
and
the
alveoli.3
10
"
An
investigation
by
May
et
al
found
symptoms
of
fever,
muscular
pain,
chest
tightness,
cough,
and
headache
to
be
directly
caused
by
organic
toxic
dust
and
suggested
that
this
"pulmonary
mycotoxicosis"
may
represent
a
systemic
reaction
to
inhaled
fungal
toxins.'2
Although
their
study
was
concerned
with
acute
episodes
after
exposure
to
massive
doses
of
organic
dust,
possibly
similar,
though
less
severe,
symptoms
occur
as
a
chronic
response
to
prolonged
exposure
to
low
c6ncentrations
of
fungal
toxins.
Analysis
of
the
moulds
collected
from
the
dwellings
in
our
study
is
still
proceeding
and
a
supplementary
report
on
the
relation
of
specific
moulds
to
symptoms
will
be
prepared.
Single
dwellings
in
the
study
were
found
to
be
harbouring
over
15
species
of
mould
and
probably
some
of
these
would
give
rise
to
allergenic
or
toxic
reactions,
or
both.
Emotional
symptoms
in
children
such
as
irritability
and
unhappiness
are
probably
linked
to
physical
symptoms
and
indicate
that
the
mental
health
of
children
is
also
at
risk.
Some
of
the
adults'
symptoms
are
difficult
to
explain
by
reference
to
mould,
though
aching
joints
and
nausea
could
both
be
reactions
to
fungal
toxins.
Reports
of
"bad
nerves"
are
not
surprising
where
living
areas
are
unpleasant,
children
are
sick,
and
family
life
may
be
fraught.
Backache
and
constipation
are
puzzling
phenomena
and
may
be
indirect
consequences
of
conditions
in
the
home.
Breathlessness
and
blocked
nose
may
be
more
closely
related
to
low
temperature.
Increased
blood
pressure
and
hypoxia
have
been
observed
as
reactions
to
cold
stress.'
We
have
attempted
at
all
stages
of
this
study,
which
is
probably
the
largest
of
its
kind
ever
undertaken,
to
refute
the
null
hypothesis-namely,
that
there
is
no
relation
between
housing
conditions
and
health
state.
To
that
end,
we
adopted
double-blind
interviewing
procedures,
included
a
wide
array
of
possible
con-
founding
factors,
and
used
multivariate
statistical
techniques.
Having
eliminated
(as
far
as
possible)
alternative
explanations
for
our
findings,
we
concluded
that
damp
and
mouldy
dwellings
have
direct
dele-
terious
effects
on
the
physical
and
psychological
well-
being
of
adults
and
children.
Our
confidence
in
this
conclusion
is
enhanced
in
more
positive
fashion
by
two
observations:
firstly,
the
similarity
of
these
findings
with
those
reported
in
our
earlier
study,'
especially
concerning
children's
respiratory
symptoms;
and,
secondly,
the
strong
relation
between
increasing
doses
of
adverse
housing
conditions
(dampness,
mould
growth,
and
air
spore
concentration)
and
symptoms
of
ill
health,
which
is
unlikely
to
be
the
result
of
respondent
bias.
A
considerable'
body
of
evidence
now
exists
that
supports
the
contention
that
dampness
and
mould
is
an
important
public
health
issue,
not
solely
for
its
immediate
impact
but
also
for
the
longterm
impli-
cations.
Poor
housing
conditions
in
childhood,
for
example,
are
associated
with
higher
rates
of
admission
to
hospital
and
higher
morbidity
and
mortality
in
adult
life.'4"
Hopefully,
planners,
policy
makers,
and
medical
practitioners
will
now
plan
concerted
joint
action
to
eradicate
this
unacceptable
and
needless
health
risk.
This
study
was
supported
by
grants
from
Glasgow
and
Edinburgh
district
councils
and
the
London
Research
Centre.
Many
associations
and
people
have
contributed
to
this
research.
In
particular,
we
acknowledge
advice
and
practical
assistance
from
the
Community
Health
Resource
Unit,
Glasgow;
Easthall
Residents'
Association;
Royston,
Molendinar
community
councils,
and
the
Technical
Services
Agency,
Glasgow.
1
M\artin
CJ,
Platt
SD,
Hunt
SM\.
Housing
conditions
and
ill
health.
Br
MedJ
1987;294:
1125-7.
2
Institute
of
Environmental
Health
Officers.
Mfould
fungal
spores-their
effects
on
health
and
the
control,
prevention
and
treatment
of
mould
gro(wth
in
dwellings.
London:
IEHO,
1985.
3
Gravesen
S.
Fungi
as a
cause
of
allergic
disease.
Allergy
1979;34:135-54.
4
Larsen
LS.
A
three
year
survey
of
microfungi
in
the
air
of
Copenhagen
1977-1979.
Allergy
1981;36:15-22.
5
Goldberg
DP.
The
detection
of
psychiatric
illness
by
questionnaire.
London:
Oxford
University
Press,
1972.
6
Baker
RJ,
Nelder
JA.
The
GLIM
system
manual.
Release
3.
Oxford:
Numerical
Algorithms
Group,
1978.
7
SPSS.
SI'SS-X
user's
guide.
New
York:
McGraw-Hill,
1983.
8
Strachan
DP.
Damp
housing
and
childhood
asthma:
validation
of
reporting
of
symptoms.
Br
MedJ
1988;297:1123-6.
9
Burr
ML,
Mullins
J,
Merret
TG,
Scott
NC.
Indoor
moulds
and
asthma.J
Roy
Soc
Health
1988;3:99-101.
10
Hosen
H.
Moulds
in
allergy.
Journal
ofAsthma
Research
1978;15:151-6.
11
Maunsell
K.
Sensitisation
risk
from
inhalation
of
fungal
spores.
J
Laryngol
Otol
1954;68:765-75.
12
May
JJ,
Stallones
L,
Darrow
D,
Pratt
DS.
Organic
dust
toxicity
(pulmonary
mycotoxicosis)
associated
with
silo
unloading.
Thorax
1986;41:919-23.
13
Lloyd
E.
Cold
stress
and
ischaemic
heart
disease.
Radical
Community
Medicine
1987;30:9-1
1.
14
Folmer-Anderson
T.
Persistence
of
social
and
health
problems
in
the welfare
state:
a
Danish
cohort
experience
from
1948
to
1979.
Soc
Sci
Med
1984;18:555-60.
15
Britten
N,
Davies
JMC,
Colley
JRT.
Early
respiratory
experience
and
subsequent
cough
and
peak
expiratory
flow
rate
in
36
year
old
men
and
women.
BrMedj
1987;294:1317-20.
(Accepted
29
April
1989)
1678
BMJ
VOLUME
298
24
JUNE
1989