Social inequalities in Central and Eastern Europe during societal transformation

Martin Bobak

Department of Epidemiology and Public Health

• Macro-economic changes• Deaths and births• Russian mortality crisis• Social inequalities

– Between populations

– Within populations

• Proposed explanations

Central and Eastern Europe in 1990

0

0.1

0.2

0.3

0.4

0.5

0.6

0 20 40 60 80 100 120

GDP per capita (PPP), West=100%

Gin

i co

effi

cien

t

??

CEE

West

LA

From Unicef

Poland and Czech Rep. 1989-2000 (Data from UNICEF)

0

0.1

0.2

0.3

0.4

0.5

0.6

0 20 40 60 80 100 120

GDP per capita (PPP), West=100%

Gin

i co

effi

cien

t

West

LA

Poland1989

Poland2000

Czech1989

Czech 2000

Russia 1989-2000 (Data from UNICEF)

0

0.1

0.2

0.3

0.4

0.5

0.6

0 20 40 60 80 100 120

GDP per capita (PPP), West=100%

Gin

i co

effi

cien

t

West

LA

Russia1989

Russia2000

Poland and Czech Rep. 1989-2007 (Data from UNICEF)

0

0.1

0.2

0.3

0.4

0.5

0.6

0 20 40 60 80 100 120

GDP per capita (PPP), West=100%

Gin

i co

effi

cien

t

West

LA

PL1989

PL2000

CZ1989

CZ 2000

PL2007

CZ 2007

Russia 1989-2007 (Data from UNICEF)

0

0.1

0.2

0.3

0.4

0.5

0.6

0 20 40 60 80 100 120

GDP per capita (PPP), West=100%

Gin

i co

effi

cien

t

West

LA

RU1989

RU2000

RU 2007

Real total consumption expenditure in “Central Europe” (from Billingsley)

CZSKHUPO SLHU

SKCZ

POCZ

SLPO

SK

HU

SK

HU

PO

SLCZ

SK

PO

CZ

SLHUCZ

SL

PO

SK

HU CZ

POSL

HU

SK

SKHU

SL

CZ

PO

SK

PO

HU

SL

CZ

HU

CZ

SK

PO

SL

SK

SL

CZ

HU

PO

SK

PO

SL

CZ

HU

PO

CZ

SK

HU

SL

SK

CZ

SL

HU

PO

CZ

PO

HU

SL

SK

8010

012

014

016

0re

al to

tal c

onsu

mpt

ion

expe

nditu

re, b

ase

year

=10

0 E

CE

1990 1995 2000 2005Year

Real total consumption expenditure in “Eastern Europe” (from Billingsley)

ROBU BE

RO

RULAUKKABU

LA

BUBEUKROKARU

RU

LA

KABUUKROBE

LA

BU

MO

RO

BE

ESRU

KA

UK

MO

RU

BU

LA

BEUK

ES

KA

RO

ES

MO

LA

BU

RO

RU

BEUK

LI

KA UK

LIMO

ES

LA

RU

RO

BE

BU

KA LA

RU

LI

BU

ES

KA

BE

RO

UK

MO

BU

UK

BE

KALA

LI

ES

MO

RU

RO

RU

BU

KA

MO

LI

BE

ES

LAUK

RO

LA

BU

UK

BE

KA

RO

RU

LI

ES

MO

LI

BU

RO

UK

RU

MO

LA

BE

ES

KA

BU

RU

UKKALA

LI

ROMO

ES

BEBE

LA

ES

RO

RU

LI

BU

KAUK

MO

5010

015

020

0re

al to

tal c

onsu

mpt

ion

expe

nditu

re, b

ase

year

=10

0 E

CE

1990 1995 2000 2005Year

Deaths…

Life expectancy at age 15 in Europe, men, 1970-2005.

40

45

50

55

60

65

1970 1975 1980 1985 1990 1995 2000 2005

EUCEEFSURussia

Life expectancy at age 15 in Europe, women, 1970-2005.

55

60

65

70

1970 1975 1980 1985 1990 1995 2000 2005

EUCEEFSURussia

Standardized death rates of the EU and CIS (WHO-HFA database)

600

700

800

900

1000

1100

1200

1300

1400

1500

1970 1980 1990 2000 2010

EU members before May 2004 EU members since May 2004 CIS

SDR all causes, all ages, per 100000

Several patterns of mortality trends

• Group 1: minimal increase, general decline: Generally Central Europe (Bulgaria, Czech Republic, Hungary, Poland, Romania, Slovakia and Slovenia).

• Group 2: increase but decline since mid-1990s: Generally the Baltics plus Armenia, Azerbaijan, Georgia

• Group 3: sustained increase / fluctuations:larger parts of FSU: Belarus, Kazakhstan, Kyrgyzstan, Moldova, Russia, Ukraine.

Life expectancy by type of transition

54

56

58

60

62

64

66

68

70

72

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Mal

e li

fe e

xpec

tan

cy a

t b

irth

TT1 life expectancy

TT2 life expectancy

TT3 life expectancy

…and births

Change in fertility after 1989

5

7

9

11

13

15

17

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

Calendar year

Cru

de

bir

th r

ate

Czech Poland Hungary Estonia Russia

Crude birth rate = live births / 1000 population From Unicef, TransMONEE database

Ratio of births in women 25+ vs. <25 (Billingsley 2009)

CZ, HU, PL, SK, SLO BG, EST, LAT, LIT

RU, UKR, RO, BEL CAR

0.00

0.25

0.50

0.75

1.00

0 50 100 150 200Months since birth of first child

pre-1991 1991+

By time periodKaplan-Meier Second Birth Estimates for Women

Missing 2nd births in Russia:

Billingsley 2009

Russian mortality crisis

Death rates in Russia 1980-2007both genders, per 100,000

800

1000

1200

1400

1600

180019

8019

8219

8419

8619

8819

9019

9219

9419

9619

9820

0020

0220

0420

0620

08

SDR Crude DR

Gorbatchevelected

Coup againstGorbatchev, breakup of USSR

Roublecrisis

3M extra deaths(Men et al, 2003)

Russian mortality from all causes, 35-69 yrs, rates per 100,000

0

500

1000

1500

2000

2500

3000

Men Women

1991199419982001

Russian mortality from all causes, 35-69 yrs, rate ratios

1 1

1.57

1.42

1.181.12

1.431.3

0.5

1

1.5

2

Men Women

1991199419982001

Proportional changes in age-specific mortality in Russia

Shkolnikov et al, 2001

Proportional change in mortality by cause of death

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Infe

ctio

ns

Cance

r

CVDRes

pirato

ryDig

estiv

eLive

r cirr

h.Ext

ernal

Alcohol

poi

s.

Men Women

From Men et al, 2003

Income inequalities and health of populations

[ecological]

Relation between life expectancy at birth and percentage of post tax and benefit income received by least well off 70%

of families, 1981 (Wilkinson, BMJ 1992)

72

73

74

75

76

77

78

43 44 45 46 47 48 49 50 51

% of income received

Lif

e ex

pec

tan

cy

W Germany

USA

AustraliaCanada

UK

Switzerland

Netherlands

Sweden

Norway

r=0.86, p<0.001

INCOME INEQUALITY & MORTALITY IN CANADA & USA

Ross et al; BMJ 2000

-7-6-5-4-3-2-10123

0 0.05 0.1 0.15 0.2

Change in Gini coefficient, 1989-95

Cha

nge

in li

fe e

xpec

tanc

y at

bir

th, 1

989-

95

(yea

rs)

Slo SvkCze

Pol

Hun

RomBul

Bel EstLit

Mol

Rus

CHANGE IN LIFE EXPECTANCY BY INCREASE IN INCOME INEQUALITY, 1989-95

Marmot & Bobak, BMJ 2001

CHANGE IN LE BY SOCIAL STRESS IN 12 RUSSIAN REGIONS, 1989-94

-8

-7

-6

-5

-4

-3

-2

12 14 16 18 20 22

*UNEMPLOYMENT, LABOUR TURNOVER,CHANGE IN MARRIAGE AND DIVORCE RATE

CHANGE IN MALELIFEEXPECTANCY

(Cornia 1997)

STRESS CAUSED BY UNEXPECTED SITUATIONS

Age-sex adjusted prevalence of poor self-rated health and

life expectancy at age 15 (r -0.73, p=0.005, n=13)

BGR

CZ E

SVK

HUN

POL

ROM

SLO

BLR

UKR

RUS

ESTLVA

LIT

52

54

56

58

60

62

Lif

e e

xp

ecta

ncy

at

age

15

(ye

ars

)

10 15 20 25Prevalence of poor health (%)

Bobak et al, JECH 2007

Odds ratios (95% confidence intervals) of poor health by

societal characteristics in different multi-level models

1.00 (0.93-1.09)1.02 (0.95-1.11)Ratio of 90th/10th income percentile, per 1 unit increase

1.05 (0.84-1.32)1.13 (0.92-1.39)Ratio of 80th/20th income percentile, per 1 unit increase

1.08 (0.98-1.18)1.09 (1.00-1.19)Homicide rates (per 10/100,000 increase)

1.15 (1.03-1.29)1.20 (1.08-1.33)Corruption index, per 1 unit (0 to 10)

1.12 (0.93-1.34)1.15 (0.95-1.39)Gini coefficient, per 0.1 unit increase

1.08 (0.99-1.19)1.10 (1.01-1.20)Max. economic contraction, per 10%

0.93 (0.85-1.02)0.90 (0.83-0.98)Annual growth, per 1%

0.79 (0.68-0.93)0.75 (0.65-0.86)GDP PPP, per $5000 increase

Odds ratio adj. for age, sex and

individual SES

Odds ratio adj. for age and sexSocietal characteristic, unit of effect

Bobak et al, JECH 2007

Social inequalities in health

[within populations]

Changes in mortality (SMR) in Russia by education between 1988-89 and 1993-94 (SMR in 1988-89=100; from Shkolnikov et al).

50

70

90

110

130

150

170

190

HIGHER LOWER HIGHER LOWER

88-89 93-94

+35%

+57%

+8%

+30%

MEN WOMEN

RR: 1.6 � 1.9 in men, 1.4 � 1.8 in women

Mortality by area-based deprivation in Budapest districts, 1980-83 and 1990-93

60

70

80

90

100

110

120

130

1980-83 1990-93 . 1980-83 1990-93

Ag

e-st

and

. mo

rtal

ity

per

100

,000

Most deprivedLeast deprived

RR: 1.4 1.6 1.2 1.4

Men Women

Jozan & Forster, 1999

Mortality by education in Estonia 1989 and 2000

0

500

1000

1500

2000

2500

3000

1989 2000 . 1989 2000

Ag

e-st

and

. mo

rtal

ity

per

100

,000

PrimarySecond.Univ.

RR: Prim. 1.7 2.4 1.4 2.2Sec. 1.3 1.9 1.1 1.7 Univ. 1.0 1.0 1.0 1.0

Men Women

Leinsalu et al,2003

Increase in educational differentials in mortality between 1980s and 1990s in St

Petersburg men

0

10

20

30

40

50

60

1980s 1990s

Mor

tali

ty p

er 1

000

PrimaryHigh schoolUniversity

From Plavinski et al 2003

Trends in probability of survival in men by education (relatives’ study)

0.4

0.45

0.5

0.55

0.6

0.65

0.7

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

Calendar year

45 p

20

45 p20 = probability of living to 65 yrs when aged 20 yrs

University

Less than secondary

Birth weight by maternal education, Czech Republic 1989-1996 (differences from primary educated)

19961995199419931992199119901989

Diff

eren

ce in

birt

hwei

ght i

n gr

amm

s 260

240

220

200

180

160

140

120

University

Secondary

Vocational

Changes by Education Group, not controlling for occupational class

0

0.25

0.5

0.75

1

1.25O

dd

s R

atio

low middle high

Pre-Transition Reference

Men Women

Changes by Education Group, controlling for occupational class

0.000.250.500.751.001.251.501.752.002.25

Od

ds

Rat

io

low middle high

Pre-Transition Reference

Men

Women

***

***

******

****** *** ***

Changes by Education Group, not controlling for occupational class

0

0.25

0.5

0.75

1

1.25O

dd

s R

atio

low middle high

Pre-Transition Reference

Men Women

Changes by Education Group, controlling for occupational class

0.000.250.500.751.001.251.501.752.002.25

Od

ds

Rat

io

low middle high

Pre-Transition Reference

Men

Women

Changes by Education Group, not controlling for occupational class

0

0.25

0.5

0.75

1

1.25O

dd

s R

atio

low middle high

Pre-Transition Reference

Men Women

Changes by Education Group, controlling for occupational class

0.000.250.500.751.001.251.501.752.002.25

Od

ds

Rat

io

low middle high

Pre-Transition Reference

Men

Women

***

***

******

****** *** ***

Billingsley 2009

Russia, Generations &Gender Survey, 2006

Fertility

Proposed explanations / mechanisms for effects on heath

• Artefact• Cohort effect• [Alcohol]• Other substances (incl. risk of HIV)• Diet, smoking and other risk factors• Medical care• Social disruption & psychosocial stress • Low resilience • Poor policy response

Artefact

• Hypothesis: reporting before 1990 was “doctored”, increase after 1990 not genuine

• Reconstructed Russian life tables (Shkolnikov) showed that the mortality rise was real

Cohort effect

• Hypothesis: vulnerable birth cohorts (e.g. people born in the war) reached age when they died disproportionately more often than other birth cohorts

• Some evidence on some outcome (e.g. CVD in Leningrad Siege Study or height in HAPIEE)

• But the effects too weak to explain the huge rise in mortality

Substance abuse: not only alcohol

• Tobacco – >60% of men smoke– but effects long-term and lagged, not acute

• Intravenous drug use– 1.8M (1.8%) i.v. drug users in age group 15-64

yrs

Infectious diseases

• TB: – incidence 119 and mortality 22 per 100,000 in 2005– Important contribution by prison population

• HIV– About 400,000 officially registered cases– Estimates suggest >1M HIV+ persons (prevalence

~1.2% among 15-49 year olds)– 37% (~0.7M) of i.v. drug users are HIV+ (Mathers et al.

2008)– Incidence about 5 times higher than in “old” EU

• Hepatitis (related to iv drugs)

• Immunisation-preventable diseases seem under control

Incidence of selected infectious diseases (per 100,000), 2006 (data from WHO)

0 20 40 60 80 100

PertussisMeasles

Diphteria.

HIVViral hep.

TB

Incidence per 100,000

RussiaOld EU

Other conventional risk factors(smoking, diet, obesity, hypertension,

cholesterol etc)

• Known to be common in CEE/FSU• Variable trends

– Mostly improvements in CEE

– Stable in FSU– Rise in female smoking in FSU

• But they don’t have acute effect and cannot explain the acute mortality crisis

Health care

‘Avoidable’ mortality

• Rutstein et al. “unnecessary, untimely deaths” (1976)

• Conditions from which, in the presence of timely and effective medical care, premature death should not occur• Single case of death (illness/disability): Why did it

happen?• Rate: not every single case preventable/

manageable � reduction of incidence

‘Avoidable’ mortality: examples

• immunisation, e.g. measles

• early detection, e.g. cervical cancer

• medical treatment, e.g. hypertension

• surgery, e.g. appendicitis

Types of ‘avoidable’ mortality

Treatable (or amenable) mortality• Deaths from causes sensitive to health care

(primary & hospital care, collective health interventions eg screening)• selected cancers (breast, colorectal, testes, cervix),

diabetes <50, hypertension/stroke, surgical conditions, maternal mortality, perinatal conditions etc.

Preventable mortality• Deaths from causes sensitive to public health or

inter-sectoral policies• Lung cancer, liver cirrhosis, transport injuries

What happened where medical care failed to modernise?

• Death rate from avoidable mortality in UK and Russia similar in 1965, when little could be done

• Gap began to widen in 1970s, and has continued to do so since

0

100

200

300

400

500

600

700

1965 1975 1985 1995

Year

deat

hs/1

00,0

00

Russia United Kingdom

Source: Andreev et al. 2003

Prevalence, awareness and treatment of high total cholesterol and hypertension in Novosibirsk and Czech

Republic, both sexes, 45-69 yrs, 2002-05, (%)

0

20

40

60

80

100

Russia Czech R. . Russia Czech R.

PrevalenceAwareness*Treatment*Control*

* Among positives

Cholesterol ≥ 5 mmol/l Blood pressure ≥140/90

Social stress

• Social disruption• Anomie• Distrust• Low social capital• Low social cohesion• High levels of inequalities• Low optimism / perceived control / mastery• Depression / hostility …

-7-6-5-4-3-2-10123

0 0.05 0.1 0.15 0.2

Change in Gini coefficient, 1989-95

Cha

nge

in li

fe e

xpec

tanc

y at

bir

th, 1

989-

95

(yea

rs)

Slo SvkCze

Pol

Hun

RomBul

Bel EstLit

Mol

Rus

CHANGE IN LIFE EXPECTANCY BY INCREASE IN INCOME INEQUALITY, 1989-95

Marmot & Bobak, BMJ 2001

PERCEIVED CONTROL IN NATIONAL

SAMPLES AND ALL CAUSE MORTALITY

1000

1100

1200

1300

1400

1500

1600

3 3.5 4 4.5 5

RU

LI

LA

ES

HU

PO

CZ

CONTROL (AGE-SEX ADJUSTED)

Pikhart, 2000

SDR

all

caus

es (

per

100,

000)

EVERYTHING changed, not just alcohol…

• Political changes• Fall in income and wages, rise in

deprivation (and diseases caused by deprivation)

• Increase in inequalities• Unemployment, insecurity, uncertainty• All this leading to “social stress”

Reaction to economic change, Russia 2001

9

48

10

32

Pro-marketNostalgicNegativePositively indifferent

From Rose 2001

Social disruption

• Some countries (mainly in the FSU) seem very susceptible to societal shocks

• Mortality fluctuations in Russia probably the largest ever recorded

• May be vulnerable in the future

Mass privatisation and adult male mortality

rates in Belarus and Russia

Stuckler et al. 2009

Interaction between mass privatisation and social capital

Stuckler et al 2009

Unemployment and mortality in Russia

0

1

2

3

4

Unemployment Paid in goods . Unemployment Paid in goods

Adj.1 Adj. 1 + alcohol

Overall Within 1 year

Perlman & Bobak 2009

Odds ratio for 2nd birth, Russia, post 1990

0.13

0.32 0.29

0.11

0.53

0.41

0

0.2

0.4

0.6

0.8

1

Unemployed Inter-generationDSM

Intra-generationDSM

menwomen

Billingsley 2009

TB and participation in IMF programme (Stuckler et al. 2008)

Poor policy response

• Poor management of the transition• Very rapid privatisation• Corruption• Massive loss of employment• Rapid deterioration of social and health

services• No safety nets• Concerns about fertility, not mortality

Low resilience

• Russian population always prone to large fluctuations in mortality

• Impact clearly differed by geography and SES

• Few resources available• No past pattern of coping• Potential accumulation of disadvantage of

the life course

Social structure

Social Environment

Work

HealthBehaviours

Patho-physiologicalchangesOrgan impairment

Well-beingMortalityMorbidity

Genes

Early Life

Culture

BrainNeuro-endocrine and immune

Psychological

Material factors

Multiple influences on health

Alcohol

Depression

UnemploymentJob stress

ViolenceInequalitiesLow cohesion

CorruptionExclusionMass privatisation

Deprivation

Cognition

Growingup in war(or 1950s)

Conclusions

• Rapid unexpected changes in social environment

• Increase in absolute deprivation• Rise in inequalities and relative deprivation• Rise in uncertainty / insecurity• Many resent the new regime

• Proximal mechanisms not clear, may include: alcohol, drugs, nutrition, other risk factors, medical care, direct effect of stress …???