People, Places and Policies: Making People
Places and Policies: Making
Multifamily Housing Healthy
Gary Adamkiewicz, PhD MPH
Harvard School of Public Health
Harvard School of Public Health
Delaware Healthy Homes Summit
3.12.2014
 Overcrowding
 High rates of infectious disease (epidemics)
High rates of infectious disease (epidemics)
 Cholera
 Tuberculosis
 Typhoid fever
 Poor sanitation  Fire hazards
 Poor lighting
 No ventilation
(Bellinger DC and Bellinger AM, JCI, 2006)
• Prevalence increasing worldwide
• Currently affects 5‐10% of US
• >17
17 million Americans affected
million Americans affected
• Incidence, severity increasing
• Highest in industrialized countries
• Leading chronic illness among children
• Prevalence among children rose from 3.6% (1980) to 5.8% (2005)*
• In the US, costs exceed $14 billion/yr
*(NEJM, 2006)
UCLA Center for Health Policy Research report
"Income Disparities In Asthma Burden and Care In California“ 2010
(NYT, 4/19/2003)
Ab t
Asbestos
SBS
Radon
Legionnaires
disease
ETS
NO2, CO
CO
VOCs
Formaldehyde
L d
Lead
Moisture/Mold
/
Allergens
Pesticides/
Ch i l
Chemicals
Ambient
A
bi
Pollution
fl
flame retardants
d
phthalates
phthalates
hh l
energy
PCB
PCBs
ventilation
PAH
pesticides
Genes
Health
Environment
Environment
• Diet
• Environmental exposures
• Physical activity
• Occupation
• Neighborhood
• Psychosocial stress
h
i l
• Healthcare • etc.
disparity linkages
disparity linkages
Indoor Environments
Sources
Indoor Sources
• Cooking appliances
• Tobacco smoke
• Cleaning products
Cl
i
d
• Air fresheners
• Personal care products
• Furnishings
• Pesticides
P i id
• Pollutant reservoirs
• Water sources
Outdoor Sources
Outdoor
Sources
• Traffic
• Industrial Activity
• Residential Activity
• Contaminated soil
Contaminated soil
Structure
Behavior
Physical Structure • Size/design of structure
• Age
• Size of living space
Si
f li i
• Single family vs. multifamily
• Leakage and/or air exchange • Heating systems
• Mechanical ventilation
M h i l
il i
Source use patterns
• Cooking appliance usage
• Cooking practices
• Smoking behavior
S ki b h i
• Consumer product usage
• Personal care product usage
Activity Patterns
• Time spent at home
• Interaction with sources
• Influence on air exchange
(Adamkiewicz et al., 2011))
Low SES ↑
• Combustion by-products
• Lead
• Allergens
g
• Mold
• Pesticides
• ETS
• Formaldehyde
• Some VOCs
• PBDEs
High SES ↑
Exposures associated w/single
family homes
• Radon
• VOCs from attached garages
‘Modern’ chemicals (e.g. SVOCs) ??
MANY CASES OF SOURCE GRADIENTS TOWARD LOW SES
Housing Variable
Low
High
Income
Income
Built before 1980 (%)
71.56
48.63
Area of peeling paint larger than 8 x 11 (%)
3.11
0.99
Any inside water leaks in last 12 months (%)
9.14
7.98
Neighborhood with heavy street noise/traffic (%)
28.19
16.69
Industry/factory within ½ block (%)
Industry/factory within ½ block (%)
6 90
6.90
1 74
1.74
Unit uncomfortably cold for 24+ hrs (%)
10.70
6.71
Evidence of rodents in unit (%)
17.77
16.26
Mean floor area of unit (ft2) 1524
2853
Mean occupant density (number per 1000 ft2)
2.78
1.82
,
,
g( )
Homes with cracks in floor, wall, or ceiling (%)
7.13
3.31
Homes with holes in floor (%)
1.85
0.37
(Low income = <$30k/yr; High income = >$100k/yr))
(Adamkiewicz et al 2011, using data from AHS, 1999)
Housing Variable
Low
High
Income
Income
Built before 1980 (%)
71.56
48.63
1.5
Area of peeling paint larger than 8 x 11 (%)
3.11
0.99
3.1
Any inside water leaks in last 12 months (%)
9.14
7.98
1.1
Neighborhood with heavy street noise/traffic (%)
28.19
16.69
1.7
Industry/factory within ½ block (%)
Industry/factory within ½ block (%)
6 90
6.90
1 74
1.74
40
4.0
Unit uncomfortably cold for 24+ hrs (%)
10.70
6.71
1.6
Evidence of rodents in unit (%)
17.77
16.26
1.1
Mean floor area of unit (ft2) 1524
2853
Mean occupant density (number per 1000 ft2)
2.78
1.82
Homes with cracks in floor, wall, or ceiling (%)
Homes with cracks in floor, wall, or ceiling (%)
7.13
3.31
0.5
1.5
2.2
Homes with holes in floor (%)
1.85
0.37
5.0
(Low income = <$30k/yr; High income = >$100k/yr))
Ratio
(Adamkiewicz et al 2011, using data from AHS, 1999)

Motivated by asthma concerns

Indoor exposures and risk factors





Allergens (Peters et al. 2008)
C b ti b
Combustion by‐products (Zota
d t (Z t et al. 2005)
t l 2005)
Pesticides (Julien et al. 2007)
Ventilation (Zota
e a o ( o a eet al. 2005)
a 005)
Smoking (Kraev et al. 2009)

Disease burden / vulnerability
Disease burden / vulnerability

Opportunities for intervention
 Single landlord
 Initiatives / policies / construction / re‐hab

4‐yr National Cancer Institute study (PI: G. Sorensen)
OVERALL GOALS OF STUDY OVERALL
GOALS OF STUDY
 To study social and physical determinants of cancer risk‐
related behavior among residents of low‐income housing THIS ANALYSIS
 To understand the prevalence of key environmental hazards within low‐income housing in the Boston area
ithi l
i
h i i th B t
 To understand the extent of clustering of these hazards
 To examine associations with self‐reported health among residents of these housing units





Cambridge, Chelsea, Somerville
15 bli & 5 i t l
15 public & 5 privately managed
d
Family units
Approx 40 households per
Approx. 40 households per development (n=828 subjects)
Surveys + visual inspections
y
p
Demographics
Characteristic
n (%) Age ‐ n (%)
18‐29 30‐39
40‐49
50‐59
60+
153 (19%)
218 (26%)
169 (21%)
(
)
145 (18%)
140 (17%)
Gender ‐ n (%)
Male
Female 169 (20%)
169
(20%)
659 (80%)
Race/ethnicity ‐ n (%)
Hispanic Non‐Hispanic White Non‐Hispanic Black Other Income below poverty Income
below poverty ‐ n (%)
n (%)
Yes
No
341 (41%)
93 (11%)
316 (38%)
74 (9%)
445 (58%)
327 (42%)
Characteristic
n (%) Survey language ‐ n (%)
English
Spanish
Creole
445 (54%)
221 (27%)
162 (20%)
(
)
Education ‐ n (%)
Grade school or below (<8 yrs) Some HS (9‐11.5 yrs) High School (12 yrs)
High School (12 yrs) More than high school (13+ yrs) 152 (21%)
123 (17%)
200 (27%)
200 (27%)
261 (35%)
Years living in development ‐ n (%)
0‐5 years
5‐10 years
More than 10 years
382 (48%)
184 (23%)
226 (29%)
Number of people in apartment
Mean (SD) [Range]
Mean (SD) [Range]
3.0 (1.5) [1‐13]
3.0 (1.5) [1
13]
Demographics
Characteristic
n (%) Age ‐ n (%)
18‐29 30‐39
40‐49
50‐59
60+
153 (19%)
218 (26%)
169 (21%)
(
)
145 (18%)
140 (17%)
Gender ‐ n (%)
Male
Female 169 (20%)
169
(20%)
659 (80%)
Race/ethnicity ‐ n (%)
Hispanic Non‐Hispanic White Non‐Hispanic Black Other Income below poverty Income
below poverty ‐ n (%)
n (%)
Yes
No
341 (41%)
93 (11%)
316 (38%)
74 (9%)
445 (58%)
327 (42%)
Characteristic
n (%) Survey language ‐ n (%)
English
Spanish
Creole
445 (54%)
221 (27%)
162 (20%)
(
)
Education ‐ n (%)
Grade school or below (<8 yrs) Some HS (9‐11.5 yrs) High School (12 yrs)
High School (12 yrs) More than high school (13+ yrs) 152 (21%)
123 (17%)
200 (27%)
200 (27%)
261 (35%)
Years living in development ‐ n (%)
0‐5 years
5‐10 years
More than 10 years
382 (48%)
184 (23%)
226 (29%)
Number of people in apartment
Mean (SD) [Range]
Mean (SD) [Range]
3.0 (1.5) [1‐13]
3.0 (1.5) [1
13]
Characteristic
Environmental indices
Environmental
indices [n with problem (%)]
[n with problem (%)]
CHEMICAL
Frequent use of pesticides or air fresheners in the homes
MOLD
Vi ibl
Visible mold or mold treatment reported by resident or visible
ld
ld t t
t
t db
id t
i ibl
mold noted during inspection
SECONDHAND SMOKE
Any reported smoking within the home y p
g
PESTS
Frequent sightings of mice, cockroaches or rats
COMBUSTION BY‐PRODUCTS
Gas stove without mechanical kitchen exhaust or reported use of
Gas stove without mechanical kitchen exhaust or reported use of gas stove to heat apartment
INADEQUATE VENTILATION
Inadequate kitchen or bathroom mechanical ventilation
Summed index
Mean (SD) [Range]
n (%)
663 (81%)
358 (43%)
172 (22%)
443 (54%)
415 (52%)
368 (48%)
2.9 (1.3) [0‐6]
Pe
ercent
40
30
20
0
10
0
0
1
2
3
4
5
6
Summed Index
These hazards were likely to be clustered
46% of homes had four or more hazards
of homes had four or more hazards
Site
Environmental Indicators
Household
Site
Current chemical usage
Current
chemical usage
Mold
Current smoking
Pests
Combustion by‐products
q
Inadequate ventilation
Household
Site
p < 0.001
Important to think about structural/systemic issues
Current chemical usage
Current
chemical usage
Mold
Current smoking
Pests
Combustion by‐products
q
Inadequate ventilation
Household
Important to think about Important
to think about
occupant behavior
Site
Environmental Indicators
Household
Environmental Indicators
0
Self‐reported health
*
*
Odds Ratio = 2.15 (1.4,3.3)
p 0.001
p<0.001
Odds of self‐reporting health as poor to fair, p
,
with good to excellent as control
All associations are adjusted for site, age, race/ethnicity, poverty status, survey language, education, having All
associations are adjusted for site age race/ethnicity poverty status survey language education having
a child under 5 in the household, having an adult over 65 in household, tenure in apartment, gender and 'ever smoked’. *p<0.05
(Adamkiewicz et al., 2013))
p for trend = 0 008
p for trend = 0.008
(Adamkiewicz et al., 2013))
Environmental Tobacco Smoke (ETS)
• Secondhand smoke contains more than 4,800 chemicals, of which at least 60 are known to cause cancer • There is no safe level of secondhand smoke exposure. Period.
• Health effects:
 lung infections
 ear problems
 SIDS
 asthma
 lung cancer (20‐30% increase)
 coronary heart disease (25
coronary heart disease (25‐30%
30% increase)
• Some groups may be more susceptible
•
Smoke travels between units through air ducts, cracks in the walls and floors, i d
k i h
ll
d fl
elevator shafts, and electrical lines; and,
•
Ventilation systems and engineering Ventilation
systems and engineering
fixes do not eliminate secondhand smoke; so,
•
Eliminating indoor smoking in multiunit housing is the only way to completely protect people from exposure
protect people from exposure 

All buildings have unintended air pathways
In multifamily, this occurs
 between units
 between units and common areas
between units and common areas
 between units and outdoors

Common pathways:
 pipe penetrations
pp p
 doors/windows

•
There is no effective way to eliminate all
paths for SHS movement
Eliminating indoor smoking in multiunit housing is the only Eli
i ti i d
ki i
lti it h i i th
l
way to completely protect occupants from exposure

Recent study: NCI‐funded study of housing developments in Cambridge
housing developments in Cambridge, Somerville and Chelsea Non‐smoking units
Major findings
‐ Measurable levels of nicotine in all but one non‐smoking unit
‐ In some units, equivalent to almost a cigarette per day
cigarette per day
‐ Residents who reported smelling cigarettes smoke from other units frequently had higher
levels of nicotine in the air

Recent study: NCI‐funded study of housing developments in Cambridge
housing developments in Cambridge, Somerville and Chelsea How often do
you smell cigarette ll i
tt
smoke from other units?
More than
Once/week
Non‐smoking units
Once/week Once/week
or less These results are consistent with
consistent with SHS moving between units or between common areas and units
and units
smokers
k
 gradient by odor report
nonsmokers
(Kraev et al. 2009)
 On Oct 1, 2012, Boston became the largest city in the country to make
largest city in the country to make it’s public housing smoke‐free
Bill McGonagle




NIH RO1 with MGH
Non smoking families
Non‐smoking families
Boston and Cambridge (control)
Surveys, nicotine, salivary cotinine (in‐home)
 Common area (nicotine/PM2.5)
 Common‐area measurements show
measurements show differences by resident characteristics and by smoking policy
•
•
•
Pesticide and other Chemical Exposures
• widespread use
• few studies of residential exposure
Wide range of potential health effects
• EDC endpoints
• cancer
• respiratory
p
y ((e.g.,
g , asthma))
• fetal development
• ADHD
• etc.
etc
Few studies addressing
• disparities
• determinants of exposure
From Weschler and Nazaroff (2008)
Organochlorines
Organophosphates
Pyrethroids
???
Banned:
DDT (1972)
DDT (1972)
chlordane (1998)
Banned:
diazinon (2001)
chlorpyrifos (2000)
Current ingredients:
permethrin
cypermethrin
etc.
• Recruitment in 2 communities
Roxbury, MA
• Roxbury, MA
• Gadsden County, FL
• Conducted approximately 200 home visits
• Survey
S
• Visual inspection
• Sampled house dust
• Lab analysis (dust samples)
• Pesticides (49 analytes)
• Other chemicals
Other chemicals
Median concentrations in house dust (ng/g)
Compound
Diazinon
Piperonyl butoxide
cis‐Permethrin
trans‐Permethrin
Cypermethrin
Class
Organophosphate
Synergist
Pyrethroid
Pyrethroid
Pyrethroid
Public Housing (n=10)
27
1720
2590
5090
3250
Other Housing (n=90)
10
180
308
731
90
p value 0.04
0.04
0.01
0.01
<0.001
Elevated levels in public housing households.
Possible dri ers
Possible drivers:
‐ infestion
‐ legacy
‐ pest control practices
BANNED
ON MARKET
DDT
chlorpyrifos
1972
2001
permethrin
INTEGRATED PEST MANAGEMENT

Focuses on what pests need for survival
◦
◦
◦
◦

Food
W t
Water
Shelter
Access
Minimizes use of ‘high exposure’ formulations
◦ Foggers, aerosols, Foggers aerosols ‘street’
street pesticides
pesticides

Uses ‘low exposure’ formulations
◦ Baits, gels, traps
Baits, gels, traps

Encouraged by HUD

CBH Study 
 Principal Investigator: Alex Lu
 Committee for Boston Public Housing
 Boston Housing Authority


Focused on exposure
Focused
on exposure to children to children
at site transitioning to IPM
IPMIIS (Integrated Pest Management Intervention I t it St d )
Intensity Study)




PI: Snehal Shah (BPHC)
Boston Public Health Commission
Committee for Boston Public Housing
Boston Housing Authority
Boston Housing Authority

Focused on understanding how IPM intensity is related to quality of life, mental health, exposures
, p

Funding: U.S. Department of Housing and Urban Development.
Funding: NIEHS 1R21ES017948‐01
Funding: NIEHS 1R21ES017948
01
 Housing represents a significant portion of energy usage globally
i
f
l b ll
Co
ontaminan
nt levels
En
nergy usagge / costs
 We need a better understanding of the tradeoffs between health and energy as we evaluate housing‐based interventions
g
Ventilation
 Need models which can evaluate interventions via
evaluate interventions via relevant metrics:
 health
 energy
 cost
energy data ‐ http://www.buildingscience.com/
existing
weatherize
renovate
‘green’
new
build
‘green’
The BRIGHT Study
Boston Residential Investigation on Boston
Residential Investigation on
Green and Healthy Transitions
Results – all households
Old Colony
Cathedral
Washington Beech
h
h
Results – all households
Old Colony
Conventional
Green
How do these transitions affect:
 Comfort?
 Satisfaction?
 Environmental exposures?
 Health?
 Energy usage?
Goal: 400+ surveys + sampling study surveys
2012 – 200+
Summer ‐100 Winter‐100
2013 – 200+
Summer ‐100 Winter ‐100
Conventional
Conventional
Green
Green
Conventional
Green
Sore or dry throat
Burning/itching eyes
Ski
Skin rashes/problems
h / bl
Green
Conventional
Inner ear infection
Sneezing attacks
Wheezing
Blurred vision
Breathing problems
Nosebleeds
Tired more than usual
Coughing
Nausea
Headaches
Dizziness
0
10
20
30
40
50
60
Percent reporting symptom in past 30 days
70
Sore or dry throat
Burning/itching eyes
Ski
Skin rashes/problems
h / bl
Green
Conventional
Inner ear infection
Sneezing attacks
Wheezing
Blurred vision
Breathing problems
Nosebleeds
*p<0.05
Tired more than usual
Coughing
Nausea
Headaches
Dizziness
0
10
20
30
40
50
60
Percent reporting symptom in past 30 days
70
35
Conventional
30
Green
25
*p<0.05
p
20
15
10
5
0
0
1
2
3
4
5
Number of Symptoms in Past Month
6
>6
6
1000
PM2.5
10
Y
Year
2013
1
2012
Formaldehyde
100
NO2 (UG/M3)
10
1
2012
NO2
100
PM2.5 (UG/M3)
P
100
1000
Formaldehyde
e (UG/M3)
1000
10
Y
Year
2013
1
2012
Y
Year
2013
Control
C
t l
n GM (GSD) n
Green
G
GM (GSD)
P
Percent Change
t Ch
Environmental Measures
PM2.5 (µg/m3)
NO2 (µg/m3)
Formaldehyde
y (µg/
(µg/m3)
Nicotine (µg/m3)
42
43
41
4
42
15.1 (2.3)
63.2 (1.8)
9.4 (2.1)
9
4( )
0.11 (5.4)
18
18
18
18
8.9 (1.8)
21.4 (1.2)
12.1 (1.8)
( )
0.02 (5.1)
‐ 41.1 %
‐ 66.1 %
+ 29.7 %
97
‐ 81.8 %
Health Measures
Symptom Summation
43
3.8 (2.5)
18
1.9 (2.1)
‐ 50.0 %
Percent Diffe
erence in Green Ho
omes vs. Control Homes
100
80
60
40
20
0
‐20
‐40
‐60
‐80
‐100
PM2.5*
NO2*
Formaldehyde
Nicotine *
Median CO2
AER† Figure 1. Percent difference in concentration (PM2.5, NO2, Formaldehyde, Nicotine, CO2) and changes/hour (AER) in green homes compared to control homes
control homes. Values calculated using multivariate longitudinal linear mixed effects models, adjusted for year and mean ambient temperature. Error bars indicate 95% confidence interval.
† AER only collected in second year of study. Multivariate model not adjusted for year. Indoor Environments
Sources
Indoor Sources
• Cooking appliances
• Tobacco smoke
• Cleaning products
Cl
i
d
• Air fresheners
• Personal care products
• Furnishings
• Pesticides
P i id
• Pollutant reservoirs
• Water sources
Outdoor Sources
Outdoor
Sources
• Traffic
• Industrial Activity
• Residential Activity
• Contaminated soil
Contaminated soil
Structure
Behavior
Physical Structure • Size/design of structure
• Age
• Size of living space
Si
f li i
• Single family vs. multifamily
• Leakage and/or air exchange • Heating systems
• Mechanical ventilation
M h i l
il i
Source use patterns
• Cooking appliance usage
• Cooking practices
• Smoking behavior
S ki b h i
• Consumer product usage
• Personal care product usage
Activity Patterns
• Time spent at home
• Interaction with sources
• Influence on air exchange
(Adamkiewicz et al., 2011))
Indoor Environments
Sources
Structure
Behavior
smoking
pesticides
pesticides
PBDEs
combustion by-prod
by prod
pesticides
PBDEs
combustion by-prod
by prod
combustion by-prod
by prod
Programs, policies, systems
H
Households, families
h ld f ili
Keys to success:
• Address root causes
• Understand systems
• Every encounter
Every encounter is an is an
opportunity
• p
people
p
• homes
• Need new partnerships
Remember – the home is a system
• Weatherization can improve IAQ, Weatherization can improve IAQ
but can also lead to problems
• Renovations can change g
dynamics
• Air sealing can lead to increased humidity, increased environmental exposures, backdrafting, and
exposures, backdrafting, and changes in airflow patterns which decrease IAQ
• BPHC’s analysis of BRFSS data showed decline in
data showed decline in current asthma symptoms from 2006‐2010 for BHA residents
• No decline seen for other housing types
g yp
• Differences remained after adjustments for age, gender ethnicity smoking
gender, ethnicity, smoking habits, health insurance, and whether foreign‐‐born
1/1/2013



Jack Spengler
Jon Levy
Jon Levy
Glorian Sorensen




Jose Vallarino
Meryl Colton
Marty Alvarez‐Reeves
Marty Alvarez
Reeves
Joan Arnold




Patricia Fabian
Patricia
Fabian
Sophia Qiu
Raphael Arku
Piers MacNaughton
Piers MacNaughton


Boston Housing Authority
Committee for Boston
Committee for Boston Public Housing




Tina Wang
Tina
Wan
Oscar Zarate
Kathleen Attfield
Robin Dodson
Robin Dodson
HIC Study
 Principal Investigator: Glorian Sorensen
 Marty Alvarez‐Reeves
M
Al
R
 Reggie Tucker‐Seeley
 Brittany Bricen
 Lorraine Wallace
 Ruth Lederman
 Funding: National Cancer Institute (R01 CA111310‐01A1). 
BRIGHT Study
 Principal Investigator: G. Adamkiewicz
 Meryl Colton
M l C lt
 Marty Alvarez‐Reeves
 Jose Vallarino
 John Kane (BHA)
 Mae Bennett‐Fripp (CBPH)

Funding: U.S. Department of Housing and Urban Development (MALHH0229 10)
Urban Development (MALHH0229‐10). 
FreshAir Study
 Principal Investigator: Principal Investigator Doug Levy
Doug Levy
 Raphael Arku
 Piers MacNaughton
 Jose Vallarino

Funding: NHLBI (RO1HL112212‐01A1). 
thanks
h k