Nicotine and smoking: insights from the Health Survey for England Martin Jarvis CRUK Health Behaviour Unit University College London.

Nicotine and smoking: Nicotine and smoking: insights from the Health insights from the Health

Survey for EnglandSurvey for England

Martin JarvisMartin Jarvis

CRUK Health Behaviour UnitCRUK Health Behaviour Unit

University College LondonUniversity College London

Why the HSE?Why the HSE?

General Household Survey is official General Household Survey is official source for national prevalence and source for national prevalence and has excellent coverage of socio-has excellent coverage of socio-economic factors associated with economic factors associated with smoking. So why bother with HSE?smoking. So why bother with HSE?

Crucial added ingredient is cotinine, Crucial added ingredient is cotinine, a quantitative measure of recent a quantitative measure of recent smoke inhalationsmoke inhalation

The nicotine model: scope The nicotine model: scope and caveatsand caveats

Nicotine does not drive smoking Nicotine does not drive smoking behaviour in a simple way to the exclusion behaviour in a simple way to the exclusion of other influencesof other influences

Nicotine-seeking modulated by individual, Nicotine-seeking modulated by individual, family, social and societal influencesfamily, social and societal influences

Understanding how nicotine intakes vary Understanding how nicotine intakes vary with these factors is the aimwith these factors is the aim

Implications especially forImplications especially for Product modification and harm reductionProduct modification and harm reduction Cessation treatmentsCessation treatments

Issues to consider…..…Issues to consider…..…

PersonPerson Reliability of self-reportReliability of self-report Preferred dosePreferred dose ConsumptionConsumption Occasional smokingOccasional smoking

ProductProduct PopulationPopulation Passive smokingPassive smoking

Cotinine as a measure of Cotinine as a measure of nicotine intake and nicotine intake and

dependencedependence Main metabolite of nicotine, with t Main metabolite of nicotine, with t ½ ½

~16-20 ~16-20 hourshours

Quantitative measure of nicotine intakeQuantitative measure of nicotine intake Spot sample may give reasonable indication of Spot sample may give reasonable indication of

steady state – hence of usual daily intakesteady state – hence of usual daily intake Conversion factor: saliva cotinine 100ng/ml = Conversion factor: saliva cotinine 100ng/ml =

~6.7mg nicotine per day. Individual variation in ~6.7mg nicotine per day. Individual variation in nicotine metabolism, but estimates should be nicotine metabolism, but estimates should be reasonably accurate when averaged across reasonably accurate when averaged across groupsgroups

Cotinine is a good indicator of the extent of Cotinine is a good indicator of the extent of nicotine dependence nicotine dependence

What has cotinine got to What has cotinine got to offer?offer?

Objective check on prevalence and cessationObjective check on prevalence and cessation Objective characterisation of extent of recent Objective characterisation of extent of recent

nicotine usenicotine use Indicator of extent of nicotine dependenceIndicator of extent of nicotine dependence Measuring dependence in relation to:Measuring dependence in relation to:

Smoking historySmoking history Occasional smokingOccasional smoking Socio-economic statusSocio-economic status Product smokedProduct smoked

Health Survey for Health Survey for EnglandEngland

Representative sample nationally aged Representative sample nationally aged 16+16+

Home interview + nurse visit to take Home interview + nurse visit to take biological measurementsbiological measurements

Data collected on demographics, smoking Data collected on demographics, smoking history at interview, saliva cotinine (gas history at interview, saliva cotinine (gas chromatography) at nurse visitchromatography) at nurse visit

Sample with cotinine:Sample with cotinine: 1998 10,532 smokers 2,6891998 10,532 smokers 2,689 2001 9,451 smokers 2,1952001 9,451 smokers 2,195

.1 1 10 100 1000

Saliva cotinine (ng/ml)

0

200

400

600

800

1000

Nu

mb

er o

f re

spo

nd

ents

.1 1 10 100 >7


0

200

400

600

800

1000

Nu

mb

er o

f re

spo

nd

ents

.1 1 10 100 1000


0

200

400

600

800

1000

Nu

mb

er

of

res

po

nd

en

ts

HSE 1998 & 2001 combined

Reported nonsmokers- no passive

exposure

Reported nonsmokers- exposed toPassive smoking

Self-reported smokers

Self-reported smoking habits and cotinine concentrations

Health Survey for England 1998 & 2001 combined

Accuracy of self-reported Accuracy of self-reported smoking status by age in smoking status by age in

population surveyspopulation surveysAgeAge Never Never

smokerssmokersEx-smokersEx-smokers

% smoking by cotinine % smoking by cotinine (>14ng/ml)(>14ng/ml)

16-2416-24 7.37.3 20.220.225-3425-34 1.51.5 9.69.635-4435-44 1.61.6 6.36.345-5445-54 0.80.8 4.24.255-6455-64 0.50.5 4.04.065+65+ 1.11.1 3.03.0

All agesAll ages 2.02.0 4.94.9HSE 1998 & 2001 combined

Self-report: no smoking, no use of nicotine replacement products

How accurate are claims of How accurate are claims of quitting in population quitting in population

surveys?surveys?Reported timeReported time

since quittingsince quittingnn % smoking% smoking

by cotinineby cotinineMean Mean

cotininecotinine

(ng/ml)(ng/ml)

< 6 months< 6 months 8989 15.715.7 1361366-12 months6-12 months 150150 24.724.7 1721721-2 years1-2 years 152152 17.817.8 2042042-5 years2-5 years 465465 7.77.7 1111115-9 years5-9 years 662662 7.17.1 16216210-19 years10-19 years 13161316 3.83.8 18318320+ years20+ years 20732073 1.51.5 135135All quittersAll quitters 64076407 4.94.9 160160


Nicotine intake and Nicotine intake and questionnaire indicators questionnaire indicators

of dependenceof dependence

Relation of nicotine intake Relation of nicotine intake to time to first cigarette in to time to first cigarette in

smokerssmokers Fine-grained and Fine-grained and

continuous continuous relationship. relationship.

Distinction Distinction between those between those who light up who light up within 5 min and within 5 min and those who wait those who wait for 5-15 minfor 5-15 minHSE 1998 & 2001 combined

50

100

150

200

250

300

350

400

450

Sal

iva

coti

nin

e (n

g/m

l)

SALIVA COTININE BY TIME TO FIRSTCIGARETTE IN SMOKERS: HSE 2001

Mean ± 95% CI

TIME TO FIRST CIGARETTE OF THE DAY

Less than

5 minutes

5-15

minutes

15-30

minutes

30 min

to 1 hr

1 to 2

hours

More than

2 hours

n=578n=274n=328n=332n=375n=335

"How soon after waking do you smoke

you first cigarette of the day?"

Relation of nicotine intake Relation of nicotine intake to time to first cigarette in to time to first cigarette in

smokerssmokers Relationship of Relationship of

cotinine to time cotinine to time to first cigarette to first cigarette is largely is largely independent of independent of cigarette cigarette consumptionconsumption


50

100

150

200

250

300

350

400

450

Sal

iva

coti

nin

e (n

g/m

l)

SALIVA COTININE BY TIME TO FIRSTCIGARETTE IN SMOKERS: HSE 2001

UnivariateAdjusted forcigarette consumption Mean ± 95% CI

TIME TO FIRST CIGARETTE OF THE DAY

Less than

5 minutes

5-15

minutes

15-30

minutes

30 min

to 1 hr

1 to 2

hours

More than

2 hours

n=578n=274n=328n=332n=375n=335

"How soon after waking do you smoke

you first cigarette of the day?"

Cotinine by questionnaire Cotinine by questionnaire measure of dependencemeasure of dependence

Dependence Dependence scale: scale: Time to first Time to first

cigarette (1-6)cigarette (1-6) Perceived Perceived

difficulty of difficulty of abstaining for a abstaining for a whole day (0-3)whole day (0-3)

Cigarette Cigarette consumption (0-3)consumption (0-3)


1 2 3 4 5 6 7 8 9 10 11 120

100

200

300

400

500

Sal

iva

coti

nin

e (n

g/m

l)

SALIVA COTININE BY DEPENDENCE INSMOKERS IN HSE 2001

Mean ± 95% CI

DEPENDENCE SCALE SCORE

N = 224 182 66241210207179130140159229 142

Variation in nicotine Variation in nicotine intake by cigarette intake by cigarette

consumptionconsumption

Relation of cotinine to Relation of cotinine to cigarette consumption in cigarette consumption in

smokerssmokers

On average, On average, cotinine cotinine increases increases linearly with linearly with consumption consumption only up to 15-20 only up to 15-20 cigarettes a daycigarettes a day


0 1-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45+

100

200

300

400

500

Sal

iva

coti

nin

e (n

g/m

l)

SALIVA COTININE BY CIGARETTE CONSUMPTIONHEALTH SURVEY FOR ENGLAND 1998 & 2001

Mean ± 95% CI

cigarette consumption

Wide variation in preferred Wide variation in preferred nicotine intake between nicotine intake between

smokerssmokers Although on average Although on average

nicotine intake nicotine intake increases with increases with cigarette cigarette consumption, at any consumption, at any level of consumption level of consumption there is extremely there is extremely wide variation in wide variation in intake between intake between smokerssmokers

10 20 30 40

Usual daily cigarette consumption

0

100

200

300

400

500

600

700

800

900

Sal

iva

coti

nin

e (n

g/m

l)5

Health Survey for England 1998 & 2001

Cigarette consumption and nicotine intake: do light or occasional smokers

take in less per cigarette? On average, smokers On average, smokers

take in about 1 mg take in about 1 mg nicotine per cigarettenicotine per cigarette

Lighter smokers take Lighter smokers take in more per cigarette in more per cigarette than do heavierthan do heavier

Little evidence for Little evidence for take-it-or–leave-it take-it-or–leave-it smokers. smokers.

Based on equivalence 100 ng/ml saliva cotinine = 6.7mg nicotine per day(Benowitz)HSE 1998 & 2001 combined

0 1-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45+0

1

2

3

4

5

6

nic

oti

ne

inta

ke p

er c

igar

ette

sm

oke

d (

mg ESTIMATED NICOTINE INTAKE PER CIGARETTEBY CIGARETTE CONSUMPTION: HSE 1998 & 2001

Mean ± 95% CI

cigarette consumption

Product smoked and Product smoked and nicotine intakenicotine intake

Low yield cigarettes and Low yield cigarettes and smoke intakesmoke intake

Nicotine intake is Nicotine intake is largely independent largely independent of machine-smoked of machine-smoked nicotine deliverynicotine delivery

Since tar and nicotine Since tar and nicotine are highly correlated are highly correlated in smoke, tar in smoke, tar exposure is unlikely exposure is unlikely to differ between to differ between smokers of low- and smokers of low- and high-yield cigaretteshigh-yield cigarettes


0 - .1 .1- .2- .4- .5- .6- .7- .8- .9- 1.0 +

Cigarette nicotine yield (mg)

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

Nic

oti

ne

in

tak

e p

er

cig

are

tte

(m

g)

Predicted and actual nicotine intakes per cigarettesmoked by nominal nicotine yield of usual brand

Actual Predicted

Health Survey for England 1998 & 2001 combined


Nicotine yields Nicotine yields have declined by have declined by 40% since 1980, 40% since 1980, but have but have smokers’ intakes smokers’ intakes reduced?reduced?

British Regional Heart Study baseline 1978-80

.

.

.

.

BRHS '78-80 HSE 1993 HSE 1994 SHS 1995 HSE 1996 HSE 1998 HSE 2000 HSE 2001

0.3

0.5

0.7

0.9

1.1

1.3

1.5

Cig

are

tte

nic

oti

ne

yie

ld (

mg

)

CIGARETTE NICOTINE YIELDS

FROM 1978 TO 2001


On average, nicotine On average, nicotine intakes in middle-aged intakes in middle-aged men have shown no men have shown no decline as tar/nicotine decline as tar/nicotine yields have declined by yields have declined by 40%40%

Consistent with complete Consistent with complete compensation for nicotinecompensation for nicotine

Indicates effective Indicates effective product modification will product modification will need to satisfy nicotine need to satisfy nicotine needs while reducing needs while reducing toxin exposuretoxin exposure

British Regional Heart Study baseline 1978-80, men aged 40-591990s data from HSE, age-matched to BRHS

.

.

.

.

BRHS '78-80 HSE 1993 HSE 1994 SHS 1995 HSE 1996 HSE 1998 HSE 2000 HSE 2001

0

50

100

150

200

250

300

350

400

Pla

sma

coti

nin

e (n

g/m

l)

0.3

0.5

0.7

0.9

1.1

1.3

1.5

Cig

are

tte

nic

oti

ne

yie

ld (

mg

)

CIGARETTE NICOTINE YIELD AND PLASMA COTININE

IN SURVEYS FROM 1978 TO 2001

mean ± 95%CI

Socio-economic variation Socio-economic variation in nicotine intakein nicotine intake

Cotinine in smokers by Cotinine in smokers by socio-economic statussocio-economic status

There is a marked There is a marked gradient in nicotine gradient in nicotine intake by socio-intake by socio-economic statuseconomic status

Smokers in the Smokers in the most deprived most deprived category take in on category take in on average about one-average about one-third more nicotinethird more nicotine


0 1 2 3 4

200

250

300

350

Sal

iva

coti

nin

e (n

g/m

l)

Saliva cotinine by in cigarette smokers by socio-economic

deprivation category: HSE 1998 & 2001

DEPRIVATION SCOREdeprev.tc

Mean ± 95%CI

Cotinine and dependence in Cotinine and dependence in smokerssmokers

The socio-economic The socio-economic gradient in nicotine gradient in nicotine intake among intake among smokers is smokers is independent of the independent of the product smoked – product smoked – seen both in smokers seen both in smokers of manufactured of manufactured cigarettes and own-cigarettes and own-rollersrollers


0 1 2 3 4

200

250

300

350

400

450

Sal

iva

coti

nin

e (n

g/m

l)

Saliva cotinine by in smokers of manufactured & own-rolled

cigarettes by deprivation category: HSE 1998 & 2001

manufactured

own rolled

DEPRIVATION SCOREdeprev.tc

Mean ± 95%CI

Variation in nicotine Variation in nicotine intake across intake across populationspopulations

National differences in National differences in nicotine dependence among nicotine dependence among

smokerssmokers

At any level of At any level of socio-economic socio-economic deprivation, deprivation, Scottish Scottish smokers have smokers have higher nicotine higher nicotine intakes than intakes than English English smokerssmokers

Scottish Health Survey 1995, 1998Health Survey for England 1996, 1998

0 1 2 3 4

250

300

350

400

450

Sal

iva

coti

nin

e (n

g/m

l)

Plasma cotinine by deprivationin adult smokers:England & Scotland compare

England

Scotland

DEPRIVATION SCORE

Passive smoking: effects of Passive smoking: effects of partner smokingpartner smoking

Dose-response Dose-response relationship relationship between partner’s between partner’s smoking and smoking and cotinine in cotinine in nonsmoking adultsnonsmoking adults

On average, dose On average, dose is about 0.7% of is about 0.7% of active smokersactive smokers

0 1-4 5-9 10-14 15-19 20-24 25-29 30+0.1

1.0

4.0

Pla

sma

coti

nin

e (n

g/m

l)

COTININE IN ADULTS BY PARTNER'S SMOKING

Partner's daily cigarette consumption

Geometric mean ± 95%CI

Health Survey for England 1998

National differences in National differences in nicotine dependence among nicotine dependence among

smokerssmokers

At any level of At any level of socio-economic socio-economic deprivation, deprivation, Scottish Scottish smokers have smokers have higher nicotine higher nicotine intakes than intakes than English English smokerssmokers

Scottish Health Survey 1995, 1998Health Survey for England 1996, 1998

0 1-4 5-9 10-19 20-49 50+

Hours per week exposed to passive smoking

0.1

1.0

4.0

Sal

iva

coti

nin

e (n

g/m

l)

PASSIVE SMOKING: COTININE IN ADULTSBY HOURS OF EXPOSURE AND HOUSEHOLD SMOKING

SmokinghouseholdNonsmokinghousehold



0.5

2.0

Passive smoking: effects on Passive smoking: effects on childrenchildren

Similar dose-Similar dose-response in response in children with children with smoking parentssmoking parents

Exposure in Exposure in children is higher children is higher than in adults with than in adults with smoking partnersmoking partner

0 1-9 10-19 20-29 30-39 40-49 50+0

1

6

Sal

iva

coti

nin

e (n

g/m

l)

PASSIVE SMOKING: CHILDREN'S COTININEBY PARENTS' CIGARETTE CONSUMPTION

-

Parents' combined daily cigarette consumption



737 57 115 120 67 44 24n=

Passive smoking and Passive smoking and deprivationdeprivation

In both adults and In both adults and children, extent of children, extent of exposure varies by exposure varies by socio-economic socio-economic statusstatus

0 0.1-9 10-19 20-29 30-39 40-49 50+

0.1

1.0

7.0

Sal

iva

coti

nin

e (

ng

/ml)

CHILDREN'S COTININE BY PARENTS'SMOKING AND DEPRIVATION: HSE 1996

0

1

2

3+

Parents' combined cigarette consumption

Geometric mean

Deprivation

score

0 1-9 10-19 20+0

1

4

Pla

sma

coti

nin

e (n

g/m

l) l

og

sca

le

COTININE IN ADULTS BY PARTNER SMOKINGAND LEVEL OF DEPRIVATION

0

1

2

3

Partner's daily cigarette consumption


Health Survey for England

1994 & 1996 combined

Deprivation

score

10 reasons why I love 10 reasons why I love the HSEthe HSE

Large and representative sampleLarge and representative sample Household sampling frame Household sampling frame Excellent coverage of demographicsExcellent coverage of demographics Smoking module as in GHSSmoking module as in GHS Brand choice questionBrand choice question Time to first cigarette questionTime to first cigarette question Cotinine – quantitative measure of smoke Cotinine – quantitative measure of smoke

intakeintake Time seriesTime series Data can be combined across years Data can be combined across years Comparison with SHS possibleComparison with SHS possible

ConclusionsConclusions

Application of cotinine in smoking studies gives Application of cotinine in smoking studies gives the opportunity for new insights into the the opportunity for new insights into the nature and determinants of nicotine nature and determinants of nicotine dependencedependence

Cross-sectional studies can yield much Cross-sectional studies can yield much information, but longitudinal studies are also information, but longitudinal studies are also needed to understand smoking careers needed to understand smoking careers

Lack of cotinine-based surveys in many Lack of cotinine-based surveys in many countries hampers understandingcountries hampers understanding

Better understanding of the contours of Better understanding of the contours of nicotine dependence important for developing nicotine dependence important for developing more effective cessation treatments and for more effective cessation treatments and for harm reductionharm reduction

Policies that ignore or misunderstand the Policies that ignore or misunderstand the dynamics of nicotine seeking are highly likely dynamics of nicotine seeking are highly likely to failto fail

Nicotine and smoking: insights from the Health Survey for England Martin Jarvis CRUK Health Behaviour Unit University College London.

Documents

nicotine metabolism

nicotine intakes

nicotine model

caveats nicotine

relation of nicotine

smoking behaviour

societal influences

cotinine ngml