1

GenderEqualityandHealth:Areviewoftheevidence

ReportpreparedfortheVictorianHealthPromotionFoundation

DrAllisonMilner

DrTaniaKing

GenderandWomen’sHealthUnit

CentreforHealthEquity

SchoolofPopulationandGlobalHealth

UniversityofMelbourne

2

Executivesummary

Manycountrieshaveidentifiedgenderequalityasanimportanteconomicandsocial

developmental goal. The movement towards equality in political representation,

education, income, and access to health services has resulted in changes to how

everydaywomen andmenwork, cohabitate, and live. The purpose of this review

wastoassesstheimpactofgenderequalityonthehealthofbothwomenandmen

in high income countries. The rationale for this review stems from VicHealth’s

updated 2016 Action Agenda, which recognises gender as a key lever to improve

health.

Gender equality was measured in slightly different ways across the 53 studies

included in the review. However, many studies recognised gender equality in

employment (e.g., female labour force participation, pay equity), parliamentary

representation, in the family (e.g.,breadwinnerandcaring roles), and inaccess to

services.

The results of this review suggest the effects of gender equality on health varied

dependingontheoutcomestudied,sphereof life(e.g.,workandhome), leveland

time period that the association was studied. However, while there were some

mixed findings, wewould argue that the evidence generally suggests that gender

equalitywasgoodforthehealthofmalesandfemales.Insayingthis,itisnecessary

toacknowledgethedynamicnatureoftherelationshipbetweengenderequalityand

health. There is unlikely to be one linear path between a country’s movement

3

toward gender equality and the health of its population. This fact reinforces the

importance of further and in-depth investigation and evaluation of the effects of

genderequalitypoliciesandpracticesonhealthoutcomes.

TherecommendationsforVicHealtharetosupportpolicymodelsthatseekto

progressgenderequalitythrough:

• Thepromotionofequalbreadwinningandcaringroleformalesandfemales;

• Encouragelabourforceparticipationofwomen(e.g.females’engagementin

thelabourforceincludinginpositionsofpoweringovernmentandbusiness),

and;

• Theprovisionofhighqualityhealthandsocialservices, includingaffordable

childcare.

At the same time, it is important to acknowledge the need for overall shifts in

cultural gender norms for males and females. Thus, a multifaceted approach is

required to ensure that gender equality benefits the health of all Australians.

Alongsidethis,westronglybelievethere isaneedformoreresearchassessingthe

progressofgenderequalityanditsrelationshiptohealth. This is particularly

necessary given the complexity of the relationship between gender equality and

health.

4

Tableofcontents

Backgroundandintroduction.......................................................................................6

Gender,genderequity,genderequalityandhealth................................................6

TheVictorianHealthPromotionFoundation’sinterestingenderequality.............7

Summaryofreviewprocess.......................................................................................14

Howwasgenderinequalitymeasured?.....................................................................19

Whateffectdoesgenderequalityhaveonhealth?...................................................23

Workplaceinequalityandleaveentitlements.......................................................23

Withincountriesanalysis–genderequalitymeasuredwithinthefamilydomain 26

Withincountries-genderequalityinpolitics,theeconomy,socialservices,and

reproductiverights.............................................................................................29

Between countries, fertility rates, educational status of women, labour force

status..................................................................................................................30

GenderInequalityIndex.....................................................................................32

Asawhole,thesestudiessuggestthatgenderequalitymeasuredatthecountry

level was associated with an improvement in a range of health outcomes,

includingphysicalactivityandsmoking(seeboxaboveforspecifics)...............33

GenderEmpowermentMeasure........................................................................34

GlobalGenderindex...........................................................................................35

GenderDevelopmentindex(GDI)......................................................................36

Discussionrelatingtothesystematicreview.............................................................37

Mainfindings..........................................................................................................37

Limitations..............................................................................................................39

5

Relevanceofthereviewtopolicyandpractice.........................................................42

Howhavepolicymodelsofgenderequalitybeendescribed?...........................42

Whichpolicymodelwillproducethebesteffectsforhealth?...............................44

ConsiderationsforVictoriaandconclusion................................................................46

References..................................................................................................................49

Appendix....................................................................................................................76

6

Backgroundandintroduction

Gender,genderequity,genderequalityandhealth

According to theWorld Health Organization (WHO), gender refers to the socially

constructed characteristics of women and men, such as norms, roles and

relationshipsof(andbetween)groupsofwomenandmen.Thesevaryfromsociety

tosocietyandcanbechanged(WHO2015).

The WHO recognizes gender as a key structural driver of inequalities in living

conditions and, by extension, inequalities in health (CSDH 2008). ‘Gender equity’

acknowledges the different needs, preferences and interests ofmen andwomen,

seeks to eliminate inequalities and discrimination, ensure equal opportunities.

Genderequality,ontheotherhand,istheprocessofallocatingresources,programs

anddecisionmakingtoensurethatbothgendershavethesameopportunities.

TheUnitedNations(UN)recognizesthataddressinggenderequityiscrucialforsocial

change (the third Millennium Development goal explicitly encourages a move

towardsgreaterequalityofwelfareresources,rolesand lifestylesbetweenwomen

andmenincountriesaroundtheworld).However,achievinggenderequalityisstill

anongoingprocess inmanyareasof theworld. TheGlobalGenderGapReport in

144countriesshowsthatwhilethereareclosinggapsbetweenmenandwomenin

health outcomes, persistent and large economic and political gaps remain (World

EconomicForum2017).Thesegapsinkeyaspectsoflifehighlighttheinequitiesthat

continuetoexistbetweenmenandwomen.Inequitiesinaccesstoresourcessuchas

7

education, income, and political representation are recognized as having large

impactsonthehealthofgirlsandwomen,particularlyinlowincomecountries(Sen

and Östlin 2008). However, there is also some evidence that gender equality

initiatives also play a role in the health of men and women in middle and high-

incomecountries(Borrell,Palènciaetal.2014).

TheVictorianHealthPromotionFoundation’sinterestingenderequality

TheVictorianHealthPromotionFoundation(VicHealth)aimstobuildandsupporta

state in which all Victorians can enjoy improved health and wellbeing (VicHealth

2013).The2016ActionAgendadefinesfivestrategicimperativesthroughwhichthe

healthofVictorianscanbeimproved:promotinghealthyeating,encouragingregular

physical activity, preventing tobacco use, preventing harm from alcohol, and

improving mental wellbeing. Furthermore, recognizing the need to act on the

underlying determinants of health and equity, VicHealth has identified three key

themes for action that will focus efforts to improve health and health equity in

Victoria:gender,youthandcommunity.

Withinthesethemes,genderequalityisseenasacriticaldeterminantofhealthand

wellbeing. VicHealth has commissioned this evidence review to examine the

evidenceandnatureofassociationsbetweengenderequalityandpopulationhealth.

It is anticipated that the results of this review will inform policy and program

developmentinthisarea.

8

The key research question that the review addressed was: Is gender equality, as

measured by gender equality indicators, associated with better population health

outcomes?Another importantaspectof thestudy is itsability to informpractices,

policies and interventions to address gender inequality. Hence, we also seek to

examine the extent to which, policy interventions have been associated with

improvedpopulationhealthoutcomes.

9

Theoreticalframeworks

There are a plethora of theories that could be used to explain the links between

gender equality and health. The aim of this section is to provide an overview of

thosetheoriescitedinthestudiesincludedinoursystematicreview.Becauseofthis,

thetheoriesdiscussedbelowshouldinnomeansbeconsideredasanexhaustivelist

of relevantmaterials. Rather,we include these theories in order to provide some

context for the specific findings of the review. The foundation formany of these

theories stem from a range of disciplines including gender studies, psychology,

sociology, public health and economics. At a fundamental level, the theories

recognisegenderandgenderequalityaskeysocialdeterminantofhealth.

Theconvergencehypothesis

This perspective argues that increasing levels of gender equality will result in a

convergence of health outcomes by gender because of a convergence of welfare

resources, roles and stress, and health behaviours between women and men

(Backhans,Lundbergetal.2007).

Roleexpansionandstress

Role expansion refers to the shift in women’s role from being within the home

domaintoalsoencompassthepublicsphere.Femalesocialrolesarethusexpanded.

Theroleexpansionhypothesissuggeststhat individualswithseveral liferoleshave

health advantages compared to thosewith fewer roles (Thoits 1983). Somewhat

10

relatedtothis,the‘multipleattachmenthypothesis’positsthatmultiplerolesimply

multiplepointsofcommunityattachment,whichare likely toboostemotionaland

instrumentalsupports,andindoingsostrengthenwomen’shealth(Lahelma,Arber

etal.2002).However,othertheoristshavearguedthatanincreaseinsocialroleswill

increaseinrolepressure,andmayresultinroleconflictandill-health(Goode1960).

Thisisdiscussedunderthe‘multipleburdenhypothesis’below.

Reductioninprotection/reduction-inequality/institutionaladjustmenthypotheses

The reduction-in-protection hypothesis argues that gender equality reduces the

female advantage in overall health (e.g., women generally have longer life

expectancy and lower rates of mortality than men). This is similar to the

convergence hypothesis discussed above. In contrast, the reduction-inequality

hypothesisarguesthatgenderequalityincreasestheadvantage(e.g.,femalehealth

isbetteroff thanmalehealth), and the institutionaladjustmenthypothesisargues

thatgenderequality initiallyreduces(femalehealthbecomesmoresimilartomale

health) and then increases advantage (female health is better than male health)

(Pampel 2001). Essentially, this theory suggests that the relationship between

genderequalityandhealthchangesovertimeasasocietyadjuststothestructural

(e.g., policies, etc.,) and cultural factors (e.g., normsabout gender roles) thatmay

accompanygenderequality.

Relativeresourceshypothesis

Thedivisionof household labour has been conceptualised in termsof the relative

resources hypothesis. According to this theory, the division of household labour

11

withinahomeisbasedonthepowerofeachpartner:thispowerisdeterminedby

possessionof social resources, and is typicallybasedon income. The spousewith

theleastpowerwilltypicallyundertakethemoreunpleasanttasks,orthosewiththe

leastprestige(suchascleaningandchildcare).

Doinggenderhypothesis

Centraltothe‘doinggender’hypothesisisthefactthatgenderedexpectationsdirect

thewaysthatindividualsconstructgenderthroughtheirdailylives(Velde,Huijtset

al. 2013). According to this theory, in counter-normative situations, men and

womendefault to stereotypical behaviors and roles (such as doing housework for

women). This compensatory feminine or masculine behavior may be harmful to

health. Research suggests that in societies in which males are typically the

breadwinner,womenwhohavegreatereconomicpowerat an individual level are

more likely to experience poorer health because they are over-burdened with

multiplerolesandroleconflict(Velde,Huijtsetal.2013).

Multipleburdenhypothesis

According to the multiple burden hypothesis, role conflict arises when there are

competingobligationsanddemandsstemmingfrommultipleroles.Womeninpaid

employment,withdependentchildrenarelikelytoexperienceroleconflict,withthe

resultant stress contributing to poorer health outcomes (Goode 1960, Lahelma,

Arberetal.2002).

12

MethodologyWeconductedthesearchaccordingtothePRISMAapproachtosystematicreviews

(Liberati,Altmanetal.2009).Tworeviewersscreenedarticlesindependently.

Twosearchstrategieswereadopted for thissystematic reviewof the literature:1)

computer searchofdatabases, 2) reviewof reference listsof all articles retrieved.

The systematic review utilised several databases – specifically PubMed, Global

Health, PsycInfo, and Scopus. The searchwas performed inMarch 2017with no

restrictionsplacedonpublicationdate, languageorpublication type. A listof the

keywordscanbeseeninTableA1oftheAppendix.

The key inclusion criterionwas that the studymust provide ameasure of gender

equity/inequity/equality/inequality as an exposure and ameasure of health as an

outcome.Wewereprimarilyinterestedinstructuralindicatorsofgenderequality–

asmeasuredindevelopedindicatorsofthephenomena-ratherthanotherpossible

dimensionsand influencesongenderequalitysuchasnormsandvalues.Whilewe

acknowledgethatnorms,valuesandattitudesarealsoimportantaspectsofgender

equality,theydidnotariseasmeasuresofgenderequalityinthisreview.

Another inclusion criterion was that the study was conducted in high income

countries,as thesesettingsweredeemedasmostrelevanttotheVictoriacontext.

Studies were excluded if they were of qualitative design, or were theoretical or

13

descriptive publications. Studies that had been conducted in low income or

developingcountrieswereexcluded.

Wewereparticularlyinterestedinmeasuresofgenderequalitythatreferencedboth

males and females, rather than measures that referred to females only. For this

reason,studiesthatexaminedfemaleonlyindicators(withoutotherindicatorsbeing

included) such as maternity leave or reproductive rights were excluded. We also

preferredmeasures of gender equality that coveredmultiple dimensions, such as

equalityinthelabourmarket,politicalsphere,andfamilydomainratherthanthose

thatcoveredasingledimension.

14

Results

Summaryofreviewprocess

Figure1describestheprocessofthereview,includinghowtheeventualnumberof

studieswasarrivedat.Ineachofthelisteddatabases,abstracts,titlesandkeywords

weresearched,andsearchterms includedaredetailed inTable1oftheAppendix.

Using these search terms, 14155publicationswere included.After the removal of

1180 duplicates, two reviewers reviewed the titles of these publications, and

excluded12082for failuretomeetthefollowing inclusioncriteria:1)mustcontain

“genderequity/inequity/equality/inequality” in title; 2)must contain ameasureof

health in title. The full-text articles of the remaining publications were then

extractedandexaminedbytworeviewers.Studieswereretainediftheymetallof

thefollowingcriteria(aspertheabove):1)providedameasuresofgenderequality

that looked at the ratio or proportion of various social, economic, or political

measuresinmalesversusfemales;2)examinedgenderequity/equalityinrelationto

ahealthoutcome;3)wereofquantitativedesign;4)relevanttoaWesterncontext.

15

Figure1:Flowdiagramofstudyselection

Recordsidentifiedthroughdatabasesearching

(n=14155)

Additionalrecordsidentifiedthroughothersources(researchers’

libraries,referencelists)(n=24)

Recordsafterduplicatesremoved(n=12951)

Recordsscreened(n=12951)

Recordsexcluded(n=12082)

Abstractsscreened(n=869)

Abstractsexcluded(n=732)

Full-textarticlesassessedforeligibility

(n=137)Studiesincludedin

Studiesincludedinquantitativesynthesis

(n=53)

Full-textarticlesexcluded(n=84)

16

Systematicreviewsummary

Atotalof53publicationswereincludedinthesystematicreview.Table1provides

somemethodologicalandgeographiccharacteristicsofthestudies.Amoredetailed

descriptionofthestudiescanbeseeninTable2attheendofthisreport.

17

Table1:CharacteristicsofstudiesincludedVariable/Characteristic No. of

studiesGeographiclocation

International 18Europeancountries 6Sweden 15USA 10UnitedKingdom 1Spain 3

GenderEqualityIndicator(exceeds53duetomultiplemeasures) GenderInequalityIndicator(GII) 12GenderEmpowermentmeasure(GEM) 6GenderDevelopmentIndex(GDI) 2GenderGapIndex 4Genderwagegap/Incomeinequality 3Genderbasedeconomicdiscrimination 1Parentalleave 2Organisationalgendergapindex 2Women’sstatusindicator(USstatebasedmeasure) 7Composite area level measure of (some, not necessarily all offollowing): political participation, employment, legal, reproductiverights

6

Genderequalitywithinthehome/comparedtopartner 11Healthoutcome(exceeds53duetomultiplemeasures)

Mentalhealth/Depression/Psychologicaldistress 11Alcohol 6Smoking 2Suicide/suicideratio 3Infantmortality/Prenatalmortality/Under-fivemortality 1Physicalactivity/inactivity 3Humanheight 1Self-ratedhealth 8Assault/Intimatepartnerviolence/adolescentdatingviolence 7Sicknessabsences 4Obesity/BMI 2Healthylifeyears 1CVD/CHDeventsormortality 2Contraception/childbearingintentions 3Mortality/NCDmortality 1Cancer 1Lifeexpectancy 4

Population Elderly65years+ 1Adults 43Women 6Child/infant 1SingleMothers 1Adolescents 1

18

Wegroupedstudiesintofourmaingroupsbasedonthemeasureofgenderequality

(Table2):

• workandemploymentrelatedgenderequality(groupone);

• genderequalitymeasuredwithinthefamilydomain(grouptwo)

• genderequalityinpolitics,theeconomy,andsocialserviceswithincountries

(groupthree);

• gender equality in politics, the economy, and social services between

countries(groupfour).

19

Howwasgenderinequalitymeasured?

Genderequalitywasmeasuredinarangeofwaysacrossdifferentstudydesigns.

In the first group (work and employment related gender equality), a number of

studiesdevelopedoverallindicatorsofgenderequality(Sörlin,Lindholmetal.2011,

Sörlin,Hmanetal.2012,Elwér,Harrysonetal.2013).Forexample,Elwér,Harryson

et al. (2013) developed a five item scale of gender equality in a workplace,

representing the ratio of male/female number of employees, salary (ratio of

males/females), educational level of employees (ratio of males/females), and

presence of parental leave (ratio ofmales/females) and temporary parental leave

(ratio of males/females). Other studies examined the pay gap between men and

women(Aizer2010),whileothersexaminedparentalleave,asdefinedasthelength

of time a male took in relation to a female (Norström, Lindberg et al. 2012,

Johansson,Wennbergetal.2014).

Inthesecondgroupofstudies(genderequalitymeasuredwithinthefamilydomain),

indicesofgenderequality reflected theextent towhichmalesand femalesshared

parentalandbreadwinnerroles(Chandola,Kuperetal.2004,Månsdotter,Lindholm

et al. 2006, Backhans, Burström et al. 2009, Sörlin, Lindholm et al. 2011,

HammarströmandPhillips2012,Harryson,Novoetal.2012,Månsdotter,Lundberg

etal.2012,Månsdotter,Nordenmarketal.2012,Neyer,Lappegǻrdetal.2013,Eek

andAxmon2015).Forexample,Månsdotter, Lundbergetal. (2012)quantified the

extenttowhichparentshadequalresponsibilities (bothparentshadbetween40%

20

and 60% of income and caring roles), had “traditional” unequal roles (father had

>60% of income andwomen had >60% caring roles), and “untraditional” unequal

roles(mother’shad<40%caringandfathers<40%ofincome).

In the third group (within country measures), gender equality indices reflected

politics, social/economic autonomy, and reproductive rights (Yllö 1983, Yllö 1984,

Kawachi,Kennedyetal.1999,Jun,Subramanianetal.2003,Chen,Subramanianet

al. 2005, Backhans, Lundberg et al. 2007, McLaughlin, Xuan et al. 2011, Roberts

2012,WängnerudandSundell2012).Forexample,Backhans,Lundbergetal.(2007)

usedthreedimensionsandnineindicatorsofgenderequality:politicalparticipation

(proportion of women versus men in municipal councils and municipal executive

committees);divisionof labour (temporaryparental leave,proportionofpart time

workersinwomenandmen);publicsphere(proportionofmenversuswomenofall

people employed in female versus male dominated occupations; proportion of

women versus men in managerial positions); and economic resources (average

incomeandrelativepovertyinfemalesandmales).

The fourth group of studies looked at the effect of gender equality between

countries andmeasured this through combined indicators comprisingofmeasures

suchastheproportionoffemalesachievinghigher levelsofeducation,theratioof

femalestomalesinparliament,andmeasuresoffemaleparticipationinthelabour

force relative to men (Pampel 2001, Bentley and Kavanagh 2008, Tesch-Romer,

Motel-Klingebieletal.2008,ClarkandPeck2012,Ricketts2014,Chon2016).

21

Theremainingcountry levelstudiesusedtheGenderInequality Index(GII),Gender

empowermentmeasure(GEM),theGenderDevelopmentIndex(GDI)ortheGlobal

GenderIndex(GGI).TheGenderInequalityIndex(GII)isaninequalityindexthatwas

introduced in2010 in theHumanDevelopmentReportofproducedby theUnited

NationsDevelopmentprogramme(UnitedNationsDevelopmentProgramme2010).

The GII measures gender inequalities across three components of human

development: reproductive health, empowerment and economic status. The GII

superseded others, including the Gender empowerment measure (GEM) and the

GenderDevelopment(GDI).

Three indicators areused toproduceGEM:proportionof seatsheldbywomen in

nationalparliaments,percentageofwomenineconomicdecisionmakingpositions;

income share (i.e. incomes ofmales vs. females). TheGender Development Index

(GDI) emerged at the same time as GEM, and serves as the gender sensitive

complementoftheHumanDevelopment Index(HDI). WhiletheHDImeasures life

expectancy, education (adult literacy, enrolment), and incomes, theGDImeasures

gendergapsinthesedimensions.Assuch,theGDIcannotbeusedindependentlyof

theHDI.

Last, theWorldEconomicForum introduced theGlobalGender Index in2006asa

meansofmeasuring and tracking genderdisparities across a rangeof dimensions.

TheGGIcomprisesof14measuresacross4majorsubindexesordomains:economic

participation and opportunity, education attainment, health and survival, political

empowerment(WorldEconomicForum2014).

22

23

Whateffectdoesgenderequalityhaveonhealth?

The following section provides a review of studies examining the relationship

betweengenderequalityandhealth.Becauseofthelargenumberanddiverserange

ofstudiescovered,wediscusstheseunderthefollowingheadings:

• workandemploymentrelatedgenderequality(groupone);

• genderequalitymeasuredwithinthefamilydomain(grouptwo)

• genderequalityinpolitics,theeconomy,andsocialserviceswithincountries

(groupthree);

• gender equality in politics, the economy, and social services between

countries(groupfour).

Workplaceinequalityandleaveentitlements

MalesHealth FemaleshealthGenderequalityhigher(howmanystudies=7)

� psychologicaldistress✔ physicalactivity✔ lower sicknessabsence� depression andanxiety�sicknessabsence� self-ratedhealth� outpatient servicesfor mental health anddrugregister

✔ violenceagainstwomen✔ psychologicaldistress✔ depressionandanxiety� physicalactivity✔ lowersicknessabsence� outpatient services formental health and drugregister✔ self-ratedhealth

Notes:✔ beneficial effects of gender equality ✗ detrimental effects of genderequality� noeffectsofgenderequality.Numberofoutcomesabovemightnotsumtothetotalnumberofstudiesasstudiesmighthaveexaminedthesameoutcome.Reducing the wage gap was associated with lower depressive symptoms and a

reduction in violence againstwomen in two studies (Aizer 2010, Platt, Prins et al.

24

2016).AparticularlyinterestingstudybyPlatt,Prinsetal.(2016)usedasampleof

men and women that were matched on education, occupation, age, and other

factorsrelatedtowages.Amongmatchedpairsofmaleandfemaleswherewomen

reported greater income than the male, there was no significant difference in

depression, and a substantially reduced disparity in anxiety.When female income

was less thanthematchedmalecounterpart,oddsofbothdepressionandanxiety

wassignificantlyhigheramongwomenversusmen.Aizer’s(2010)studyfoundthata

reduction in the incomegapwasassociatedwitha reduction indomestic violence

acrosstheUSA.

There were three studies conducted at the workplace level (Sorlin, Ohman et al.

2011,Sörlin,Hmanetal.2012,Elwér,Harrysonetal.2013). Elwér,Harrysonetal.

(2013)foundthat,forwomen,thehighestoddsofpsychologicaldistresswerefound

in traditionally gender unequal workplaces. The lowest overall occurrence of

psychological distress was found on themost gender equal workplaces. Another

study by Sörlin, Hman et al. (2012) found that women in companies with

“completelyequal”or“quiteequal”scoresontheOrganizationalGenderGapIndex

had higher odds of reporting “good health” compared to women who perceived

theircompanyas“notequal”.Althoughnotstatisticallysignificant,thesametrends

wereobservedinmen.However,Sorlin,Ohmanetal.(2011)alsofoundhigherrates

ofsicknessabsenceatgenderequalcompanies.

Therewerealsostudiesthatexaminedworkplaceallowancessuchasparentalleave

asaformofgenderequality(Norström,Lindbergetal.2012,Johansson,Wennberg

25

etal.2014).Forexample,astudybyJohansson,Wennbergetal.(2014)studiedthe

relationship between parental leave and physical activity, finding that longer

parentalleavewasrelatedtogreaterphysicalactivityinfathers.Therewasnoeffect

onparentalleaveforwomen.

Summary:

As a whole, this group of studies suggest that greater gender equality in the

workplacewasassociatedwithbetterhealth.Although,itshouldbenotedthata

number of studies were not conclusive: this likely to be due to a number of

methodologicalfactors(e.g.,lackofstatisticalpower).

26

Within countries analysis – gender equality measured within the family

domain

MalesHealth FemaleshealthGenderequalityhigher(howmanystudies=9)

✔ childbearingintentions✔ Alcohol relatedinpatient care ormortality� mortality✗ anxietysymptoms✔ sickleave� CHD✔ SRH✔ lowerpsychologicaldistress

✔ childbearingintentions✗ Alcohol related inpatientcareormortality✗ mortality✗ sickleave✔ CHD✗ anxietysymptoms� SRH✔ stress, fatigue, physicalsymptoms✔ lower depressivesymptoms and psychologicaldistress

Notes:✔ beneficial effects of gender equality ✗ detrimental effects of genderequality� noeffectsofgenderequality.Numberofoutcomesabovemightnotsumtothetotalnumberofstudiesasstudiesmighthaveexaminedthesameoutcome.SRH=self-ratedhealth;CHD=coronaryheartdisease.

Amongthisgroup,therewereninestudiesthatexaminedgenderequalitywithinthe

family domain (Chandola, Kuper et al. 2004, Månsdotter, Lindholm et al. 2006,

Backhans,Burströmetal.2009,Sörlin, Lindholmetal.2011,Harryson,Novoetal.

2012, Månsdotter, Lundberg et al. 2012, Månsdotter, Nordenmark et al. 2012,

Neyer,Lappegǻrdetal.2013,EekandAxmon2015).

There were a number of studies that suggested that gender equality within the

familydomainwasassociatedwithpoorerhealthoutcomes(Månsdotter,Lindholm

et al. 2006, Backhans, Burström et al. 2009, Månsdotter, Lundberg et al. 2012,

27

Neyer,Lappegǻrdetal.2013).Thesestudiesclassifiedgenderequalitybytheextent

towhichmalesandfemalesincoupledrelationshipsparticipatedinhomeandpaid

work.Ingreaterdetail,womenintraditionalroles(wherewomenoccupiedmostof

thetime inthehomesphere,while theirmalepartnersspentagreateramountof

time at work) appeared to have lower risk of alcohol related inpatient care or

mortality(Månsdotter,Lundbergetal.2012),overallmortalityandsicknessabsence

(Månsdotter,Lindholmetal.2006),andlowersickleave(Backhans,Burströmetal.

2009).

Incontrast,otherstudiesthatspecificallyexaminedgenderequalitywithinthehome

find that equal relationships were associated with lower coronary heart disease

(Chandola, Kuper et al. 2004), depressive symptoms (Hammarström and Phillips

2012,Harryson,Novoetal.2012),perceivedstress,fatigue,physical/psychosomatic

symptoms, and work family conflict (Eek and Axmon 2015). This suggests that a

moreequaldivisionoftasksinthehomeresultsinarangeofpositiveoutcome.One

paper suggests that gender equality (e.g., gender balanced division of household

work)was associatedwith higher childbearing intentions (Neyer, Lappegǻrd et al.

2013),whileoneotherfounditwasassociatedwithbetterself-ratedhealth(Sörlin,

Lindholmetal.2011).

Other studies examined a persons reported change in gender ideology (e.g.,

personallyheld thoughts)over time in relation to theirparentsgenderorientation

(Månsdotter, Nordenmark et al. 2012). Results suggest that for women andmen,

non-traditional gender ideology (e.g., believing that women could be the

28

breadwinner in a relationship) in adulthoodwas associatedwith decreased risk of

anxiety symptoms.Further, thosepeoplewho,atage42,haddeparted fromtheir

traditional attitudes towardsmasculine (formen) and feminine (forwomen) roles

(reportedatage30)reportednosignificantdeclinesinmentalhealth(Månsdotter,

Nordenmark et al. 2012). Last, a parents’ gendered life was not decisive for a

person’sowngenderedlife,andadulthoodgenderpositionruledouttheimpactof

childhood gender experience on self-reported mental ill-health (Månsdotter,

Nordenmarketal.2012).

Summary:

It appears that greater equality in household tasks is related to better health.

However, it is important to take the dynamics of a person’s relationship into

account. Women who could be seen as the main income earners may also be

taking on a substantial load of housework and childcare. Thus, poorer health

among these women may reflect a lack of true equality in the division of

householdtasks,resultingingreaterstressandpoorerhealth.

29

Within countries - gender equality in politics, the economy, social services, and

reproductiverights

MalesHealth FemaleshealthGenderequalityhigher(howmanystudies=7)

✔ physicalactivity✔ loweralcoholuse✗ sicknessabsence

✔ ✗ lower violence againstwomen✔ ✗ self-ratedhealth✔ post-traumatic stressdisorderanddepression✔ loweralcoholuse✔ lowerdisability✗ sicknessabsence✔ lower mortality rates andphysicallimitations

Notes:✔ beneficial effects of gender equality ✗ detrimental effects of genderequality� noeffectsofgenderequality.Numberofoutcomesabovemightnotsumtothetotalnumberofstudiesasstudiesmighthaveexaminedthesameoutcome.

Thereweresevenstudies thatconductedwithin-countryarea investigationsof the

association between gender equality, measured in multiple domains, and health

outcomes(Yllö1983,Yllö1984,Kawachi,Kennedyetal.1999,Jun,Subramanianet

al. 2003, Chen, Subramanian et al. 2005, Backhans, Lundberg et al. 2007,

McLaughlin, Xuan et al. 2011, Roberts 2012,Wängnerud and Sundell 2012, Sanz-

Barbero, Vives-Cases et al. 2015). Themajority of studies at this level found that

gender equality (measured as political participation, reproductive rights, socio-

economic status) was associated with positive health outcomes, including lower

reportedintimatepartnerviolence(Sanz-Barbero,Vives-Casesetal.2015),self-rated

poor health (Wängnerud and Sundell 2012), mood and anxiety disorders (Chen,

Subramanian et al. 2005, McLaughlin, Xuan et al. 2011), alcohol consumption

30

(Roberts 2012), and femalemortality rates (Kawachi, Kennedy et al. 1999). There

was one study (Backhans, Lundberg et al. 2007) that found that higher gender

equality was associated with higher levels of sickness absence and lower life

expectancy.Twoolderstudies (conducted in theearly1980s)examinedstate level

equality in relation to intimate partner violence. High gender equality was

negativelycorrelatedwithviolenceagainstwomen(Yllö1983),butfurtherresearch

suggested that this was a complex relationship. Most particularly the research

suggestedthatdiscordanceorinconsistencybetweenindividualequality(balanceof

power within the relationship) and state level equality was predictive of greater

levelsofintimatepartnerviolence(Yllö1984).

Summary:

Asawhole,theevidencefromthisgroupofstudiessuggeststhatgenderequality

wasassociatedwithbetterhealthoutcomes.

Betweencountries,fertilityrates,educationalstatusofwomen,labourforcestatus

MalesHealth FemaleshealthGenderequalityhigher(howmanystudies=8)

✗ life expectancy (comparedtomen)✗ mortality (compared tomen)✗ suicide(comparedtomen)� female homicide✔ contraceptiveuse✔ SRH(comparedtomen)–culturalacceptanceofgenderequalitycritical.

31

Notes:✔ beneficial effects of gender equality ✗ detrimental effects of genderequality� noeffectsofgenderequality.Numberofoutcomesabovemightnotsumtothetotalnumberofstudiesasstudiesmighthaveexaminedthesameoutcome.

There were a number of cross-country studies that used multiple indicators of

genderequality(fertilityrates,ratiooffemalestomalesinthelabourforce,women

in parliament, policies regarding families) and health outcomes (Pampel 2001,

Bentley and Kavanagh 2008, Tesch-Romer,Motel-Klingebiel et al. 2008, Clark and

Peck 2012, Ricketts 2014, Chon 2016). These generally producedmixed findings.

Clark and Peck (2012) studied the relationship between various gender equality

measures and life expectancy between 1985 and 2005. They find that the life

expectancy in femalesdecreases (whilemale life expectancy increases) as females

begin toparticipate in“traditionalmale institutions”,suchas formalschoolingand

paidemployment,andbegintoshiftawayfromtraditionalfemaleroles,asindicated

by a decline in fertility. Another study by Chon (2016) seeks to examine the

“backlash” hypothesis, which suggests that female homicide will increase as they

question the gender order. In fact, the authors find no evidence of a significant

increaseinhomicideingenderequality.

Summary:

The mixed findings in this group of studies are likely to reflect methodological

differences.Comparedtothosementionedearlier,afocusofthesestudieswasthe

progress of gender equality over time. It should also be mentioned that the

methodologicalqualityofthesestudieswaslow.

32

GenderInequalityIndex

MalesHealth FemaleshealthGenderequalityhigher(howmanystudies=10)

✔ leisure timephysicalactivity✔ smoking✔ infant mortalityrates� SRH

✔ leisure time physicalactivity✗ smoking✔ infantmortalityrates� SRH✔ reproductivehealth✗ ✔ lowerviolence✗ ✔lifeexpectancy✔ obesity✔ height✔ smoking; alcoholconsumption; lifeexpectancy, cancer, BMI,bloodpressure

Notes:✔ beneficial effects of gender equality ✗ detrimental effects of genderequality� no effects of gender equality BMI=body mass index SRH=self-ratedhealth

TheGIIwasusedbytwelvestudies(Bond,Robertsetal.2010,Wells,Marphatiaet

al.2012,DahlinandHarkonen2013,Hassanzadeh,Moradietal.2014,KimandKim

2014,Mark2014,Brinda,Rajkumaretal.2015,Gressard,Swahnetal.2015,Balish,

Deaneretal.2016,Bilal,Beltranetal.2016,Dereuddre,VandeVeldeetal.2016,

Redding,Ruiz-Canteroetal.2016)inrelationtoarangeofhealthoutcomes

includingobesity,leisuretimephysicalactivity,lifeexpectancyandself-ratedhealth.

Of relevance to VicHealth, gender equality was also found to have beneficial

associations with obesity: higher levels of gender equality were associated with

reducedfemaleexcessofobesityinastudycomparingtheprevalenceofobesityin

68countries(Wells,Marphatiaetal.2012).AlsoofrelevancetoVicHealth,aSpanish

study examined the relationship between the gender smoking ratio (GSR) and

33

genderequality inSpain,asmeasuredusingtheGII,over50years(Bilal,Beltranet

al. 2016). A strong negative correlationwas observed between gender inequality

andtheGSRoverthe50yearstudyperiod:asgenderequalityincreased,thefemale

tomalesmokingratioincreased.Thesocialpatterningandchronologyofadoption

of cigarette smoking was apparent, with men and the highly educated the early

adopters,andwomenandthoseoflowereducationalattainmentlateradopters,and

later to cease smoking. The authors observed that sweeping social, economic and

politicalchangeshadtransformedtheculturalandsocialclimateofSpaininthelast

halfcentury.Theysuggestedthatthetobaccoindustryhadco-optedthemessageof

liberationandemancipation thataccompanied such social changes,andpresented

cigarettesmokingassymbolicofgenderequalityandemancipation(Bilal,Beltranet

al. 2016). Hassanzedeh and colleagues (2014) studied 123 countries and found

negativeassociationbetweengenderinequalityand:smoking;alcoholconsumption;

lifeexpectancy,cancer,BMI,andbloodpressure.

Summary:

As awhole, these studies suggest that gender equalitymeasured at the country

levelwasassociatedwithanimprovementinarangeofhealthoutcomes,including

physicalactivity,butwasalsoassociatedwithanincreaseinotherbehaviourssuch

assmokingandalcoholconsumption.

34

GenderEmpowermentMeasure

MalesHealth FemaleshealthGenderequalityhigher(howmanystudies=10)

✔ lowerdepression✗ suicide� alcohol

✔ lowerdepression✗ hazardousdrinking✗ suicide� alcohol

Notes:✔ beneficial effects of gender equality ✗ detrimental effects of genderequality� noeffectsofgenderequality

TheGenderEmpowermentmeasurewasusedasameasureofgenderequalityinsix

studies(Mayer2000,Shah2008,Tesch-Romer,Motel-Klingebieletal.2008,Bond,

Robertsetal.2010,Velde,Huijtsetal.2013,Bosque-Prous,Espeltetal.2015)that

aroseintheliteraturereview.

Ina studyacross25Europeancountries,genderequalitywaspositivelyassociated

with lower levels of depression among men and women, leading the authors to

conclude that, in terms of depression, both genders benefit from higher levels of

gender equality (Velde, Huijts et al. 2013). However, another study across 16

countries (also in Europe) found that greater gender equalitywas associatedwith

reduced gender differences in hazardous drinking (i.e., there was a higher

consumption of alcohol among women in countries of higher gender equality)

(Bosque-Prous,Espeltetal.2015).Thefactthattheassociationbetweenhazardous

drinkingandGEMwasstrongerinwomenthanmenledtheauthorstosuggestthat

improvedsocialconditionsforwomenmayleadthemtoadoptriskybehavioursthat

havetypicallybeenassociatedwithmales(Bosque-Prous,Espeltetal.2015).

35

Summary:

Thereweremixedfindingsamongthesmallnumberofstudiesusingthismeasure.

GlobalGenderindex

MalesHealth FemaleshealthGenderequalityhigher(howmanystudies=10)

✔ LPTI

☐ self-reportedhealth� internal or externalalcoholrelatedproblems✗ LPTI

Notes:✔ beneficial effects of gender equality ✗ detrimental effects of genderequality� noeffectsofgenderequalityLPTI=leisuretimephysicalinactivity

FourstudiesexaminedhealthoutcomesinrelationtotheGGI(Bond,Robertsetal.

2010, Grittner, Kuntsche et al. 2012, Van Tuyckom, Van de Velde et al. 2012,

Witvliet,Arahetal.2014),withnullfindingsemergingforthestudybyWitvliet,Arah

et al. (2014) andGrittner, Kuntscheet al. (2012). Another studybyVanTuyckom

and colleagues (2012) found that greater gender equality was associated with a

reduced gender gap in sedentary behaviour. However, these differences

disappeared in countries with high gender equality. The study by Bond and

colleagues (2010) used multiple measures of gender equality and is discussed

elsewhere.

Summary:

Thereweremixedfindingsamongthesmallnumberofstudiesusingthismeasure.

36

GenderDevelopmentindex(GDI)

MalesHealth FemaleshealthGenderequalityhigher(howmanystudies=10)

✔ depression

✔ depression

Notes:✔ beneficial effects of gender equality ✗ detrimental effects of genderequality� noeffectsofgenderequalityHopcroftandBradley(2007)foundthatratesofdepressionwerehigherincountries

with low gender equality, however the gender gap in depression was larger in

countries with high gender equality. The authors interpreted the results as

suggesting that the benefits of living in a country with high gender equality are

greaterformalementalhealththanforfemalementalhealth.

Summary:

Therewerebeneficialeffectsofgenderequalityonhealth.

37

Discussionrelatingtothesystematicreview

Mainfindings

The results of this review suggest the effects of gender equality on health varied

dependingontheoutcomestudied,sphereof life(e.g.,workandhome), leveland

timeperiodthattheassociationwasstudied.Whilethereweresomemixedfindings

(as discussed below), we would argue that the evidence generally suggests that

genderequalitywasgoodforthehealthofmalesandfemales.

Genderequalityintheworkplacewasgenerallyrelatedtobettermentalhealthand

overall self-rated health, particularly for women. In this context, gender equality

referred to reducing gender-based discrimination in the workplace, ensuring an

equal representationofmalesand femalesatwork, andprovidingequal access to

payandparental leave.Results fromarea level studies (examininggenderequality

withinacountry indifferentstatesor regions)alsosuggestedthatgenderequality

wasgoodforhealth.Inthesestudies,genderequalitywasmeasuredinavarietyof

ways, including political participation, labour force participation, and reproductive

rights. Similarly, between country studies (examining a countries overall level of

gender equality) suggested that gender equality was good for health. However, a

number of these studies also highlighted the likelihood of a convergence in life

expectancyforwomenandmen.

Another group of studies examined gender equality within the home. These

suggestedthatwomenandmenwhotakeonuntraditionalroles(e.g.,womenwho

38

are themain breadwinner,menwho are themain carers) hadworse health than

thosewhotakeontraditionalroles.Womeninparticularhadhigherlevelsofalcohol

use and worse mental health. However, there was also some evidence that the

context inwhichmenandwomen livedandworkmattered.Backhansetal (2009)

studyof“pioneers”and“laggards”examinedtheextenttowhichpeople’s rolesat

homeandatworkmirroredthenormsofwhatwasoccurring intheirmunicipality.

Findings suggested that fathers who had an equal relationship had better health

whentheywereincongruencewiththeoveralllevelofgenderequality.Forwomen,

being a “pioneer” (being in a couple thatwasmore equal than yourmunicipality)

conferredtheworsthealthbenefits,butresultsweresimilarforall“equal”groups.

Being “equal” in thepublic spherewas thusgenerallybad forhealth regardlessof

thelevelofgenderequalityinaperson’smunicipality.

To some extent, these findings may reflect the potentially damaging effects of

altering the gender status quo within the personal realm of the family, e.g., the

‘doinggender’ hypothesis(Velde,Huijtsetal.2013).Atthesametime,this could

reflect the effect of women’s ‘double burden’ of paid and unpaid work, where

womenhave to contendwith competingobligationsanddemands stemming from

multipleroles,andsufferpoorerhealthbecauseofthis(Goode,1960).

Area-level studies conducted within countries (i.e., where the unit of study was

municipalities or states) generally found that gender equalitywas associatedwith

better self-rated health, mental health, and lower mortality, alcohol use and

disability.Byandlarge,thesestudieswereconductedinSweden(whichisoneofthe

39

most genderequal countries in theworld)or theUnited StatesofAmerica. These

studieshighlightedthe importanceofwiderstructural reformtowomen’srights in

political and economic realms, and the power that this has to flow through to

improvetheoverallhealthofapopulation.

At a country level, studies suggested that gender equality had a mixed effect on

health.Therewasevidencethatmaleandfemalehealthoutcomesconvergedover

time (Clark and Peck 2012). This lends some support to the convergence or the

reduction-in-protection hypothesis, which suggests that that increasing levels of

genderequalitymayresult inaconvergenceofhealthoutcomesbygender,chiefly

because of a convergence of welfare resources, roles and stress, and health

behavioursbetweenwomenandmen(Backhans,Lundbergetal.2007).

Limitations

There are several limitations of this review and the studieswe includedwithin it.

Firstly,manyofthehealthoutcomemeasureswestudiedwereobtainedfromself-

reportedmeasures,which are likely to be subject to reporting bias, particularly if

reportingonsensitivebehavioursuchasmentalhealth.Further,manyoftheitems

(both outcome and exposure) were derived from single-item measures that are

unlikelytocapturemorenuancedelementsofbehaviour. Itwasnotedthat inthe

caseofdivisionofhouseholdduties, for instance,asingle itemmayfail toaccount

forbroaderdivisionofhouseholdresponsibilities (EekandAxmon2015). Inmany

studies,therewaslimitedornocontrolforconfoundingvariables,whichmayhave

40

biased estimates and in some studies there was insufficient data on some

components of gender inequality measures, and the means of handling this was

sometimes inadequate. Another limitation was the fact that many of the studies

reported here were cross-sectional and ecological. There are several limitations

associatedwithsuchdesigns.Aswithallcross-sectionalstudies,itisnotpossibleto

makecausalinference,andindeedinsomecasesreversecausationcannotberuled

out.Itisnotimplausible,forexamplethatclusteringofunhealthywomenleadsto

reducedlabourforceparticipationinsomeareas.

Theecologicalnatureofmanyofthepublicationsreviewedhereraisestheissueof

theecologicalfallacy.Theecologicalfallacycanoccurwheninferencesaremadeat

the individual level based on data aggregated at a group level (Diez Roux 2002,

MacintyreandEllaway2003).Theassociationsbetweentwovariablesmeasuredat

thegrouplevel,maydifferfromassociationsbetweentwoequivalentvariablesatan

individuallevel:theecologicalfallacyariseswhentheassociationsareassumedtobe

thesame.Forexample,itmaybethatatacountrylevel,increasedgenderequality

isassociatedwithbetterpopulationhealth. Toassume thatatan individual level,

high gender equality is associated with good health is to commit the ecological

fallacy.

It is also likely that there is some lag between exposure to gender inequality and

healthoutcomes.Whileitispossiblethatwithinacountryorstatecontext,gender

equalityatdifferenttimepointsislikelytobehighlycorrelated,thefactthatmany

studiesmeasuredgenderequalityandthehealthoutcomeatthesametimeneeds

41

tobenotedasalimitation. It isalsoworthnotingthatdifferentoutcomesmaybe

more sensitive to shifts in the social context than others, so more careful

considerationoftemporalityisneeded.

Whilenotnecessarilyalimitation,itisworthnotingthatseveralstudiesincludedin

this review were conducted in Europe, roughly around the time of the 2008

economicrecession,whenthereweresignificantshiftsinlabourforceparticipation,

as well as economic austerity measures (with likely changes in gender equality

policies),whichare likelytohavealteredthesocialmilieu,andattitudestogender

equality,inmanycountries.

Anotherimportantlimitationisthefactthatmanystudiesfocusedonfemalehealth

outcomesonly,andthereforeneglectedtoassesstheimpactofgenderequalityon

malehealth.Thismethodologicallimitationhaseffectsthatflowthroughtopolicy.

Forexample,acriticismthathasbeenmadeof“gendermainstreaming”isthat

commentaryandinvestigationoftenconflates“gender”with“women”,andthat

gendermainstreamingneglectsafocusonmen(SaundersandPeerson2013).Itis

alsoworthconsideringthatgendermainstreamingdoesnotnecessarilytranslateto

betterhealthoutcomesforwomeneither.

42

Relevanceofthereviewtopolicyandpractice

As we have discussed above, the review suggests that gender equality may have

important flow through effects to health. The following section seeks to provide

recommendations on: a) how policy and practice regarding gender equality may

benefit health; b), how to assess evidence bywhich policy and practice regarding

genderequality,mayimpacthealth.

First and foremost,wehave to acknowledge thatwhile anumberofpolicieshave

beenimplementedatvariouslevelswiththeaimofaddressingtheeffectsofgender

inequalityonhealth,therearefewstudiesthathaveevaluatedthesuccessofthese

policies(Backhans,Burströmetal.2012,Borrell,Palènciaetal.2014).

Further,keyresearchersintheareaarguethatitiscriticaltoassesstheoverallpolicy

climate of gender equality at a country level, rather than looking at policies

individually(Muntaner,Borrelletal.2010).Thisisbecauseassessmentofindividual

policiesfailstotakeintoaccounthowdifferentsocial,economicandwelfarepolicies

intersect with one another. Hence, in this section, we will discuss overall policy

regimes and approaches to gender equality at a country level, rather than

specificallyexaminingindividualpolicies.

Howhavepolicymodelsofgenderequalitybeendescribed?

Onegroupofresearchershaveconductedanumberofreviewsofgenderpoliciesin

relationtocountryregimetype(Palència,DeMoorteletal.2017).Theseresearches

43

usedaclassificationdevelopedbyKorpi,whogroupedEuropeancountriesaccording

tothecharacteristicsoftheirlabourforceparticipationpolicies(Korpi2010).

• Dual-earner/dual-carerregimesaretypifiedbyNordiccountries,whichhave

policies to support women and men’s employment and caring

responsibilities.

• Traditionalcountriesarecharacterisedbypoliciesthatsupportthetraditional

familymodel (womenascaregiversandmenasbreadwinners),and include

mostcontinentalandnorthwesternEuropeancountries.

• Market-oriented countries are those characterised by little or no policy

supportforhouseholds,andaremostlyAnglo-Saxoncountries.

Othershaveexpandedthelisttoinclude:

• Traditional southern countries: characterised by southern European

countrieswith lower female labour forceparticipation anda lackof policy

supporttofamilies,leadingtorelianceonunpaidwork,and;

• Contradictory countriesaremainly formersocialistcountrieswherethere is

both support for dual-earner families, mingled with attempts to retain

gendereddivisionofdomesticlabour.

Another study classified countries into policy regime types/clusters based on

Sainsbury’stypology(Sainsbury1999).Theseare:

• Malebreadwinner regimesare typifiedbyan ideologyofmaleascendancy,

largely based on a division of labour that is highly gendered. There is

typically favouring of the sole breadwinner through allowances and tax

credits.

44

• Earner-carer regimes are characterised by an ideology that supports equal

rights between men and women, and shared roles and responsibilities

(Sainsbury 1999), through generous parental leave, high social service,

separatetaxationandhighpensionuniversality.

Backhansetal.,expandedthistypology,andcreatedtwofurtherclassifications:

• Universal citizen which are characterised by reasonably generous and

inclusive means-tested benefits, separate taxation and high pension

universality(Backhans,Burströmetal.2012).

• The compensatory breadwinner is typified by separate taxation, and

compensatorymeasures in the pension systems (Backhans, Burströmet al.

2012).

Whichpolicymodelwillproducethebesteffectsforhealth?

Fromtheevidencewereviewabove,itisclearthattheimpactofgenderequalityon

healthiscomplexandwillchangeovertimeandthelocation.However,onthebasis

oftheevidenceabovewesuggestthatpoliciesthatsupportthedual-earner/dual-

carermodelarelikelytoresultinbetterhealthoutcomes(specifically,mentalhealth

and self-ratedhealth). Supporting this conclusion,anevaluationofdifferentpolicy

regimesinrelationtohealthoutcomesrevealedthatpoliciesthatsupportedfemale

labour forceparticipationandreducedcareburdenswereassociatedwithreduced

gendergapsintermsofhealth(Palència,DeMoorteletal.2017).Thiswasenacted

through increasedpublic services suchaschildcare,provisionofeconomicsupport

45

forfamilies,aswellasentitlementsforfathers(suchaspaternityleave)(Palència,De

Moorteletal.2017).ThisalsoalignswithareviewofthetopicbyBorrell,Palència

et al. (2014), who conducted a comprehensive review of policies and their

associationswithgenderequality.Theseresearchersalsoconcludedthatthe“dual-

earner policy model” of Scandinavian welfare states involves policies to support

women’sparticipation in theworkforce, andpolicies toencouragemoreequitable

sharing of responsibilities for unpaid work, are associated with better health for

women(Borrell,Palènciaetal.2014).

46

ConsiderationsforVictoriaandconclusion

Asweargueabove,strategiestosupportthedualearner/dualcarermodelmayhave

a number of population level health benefits. Thus, policies that support gender

equalitythrough:

• Promotionof theequal caring roleofmalesand females throughparental

leave,andflexibleworkingarrangements;

• Encouraging labour forceparticipationofwomen (e.g. femalesengagement

in the labour force including in positions of power in government and

business),and;

• High quality universal health and social services, including affordable

childcare;

arerecommendedaslikelytohavethebestpopulationhealthoutcomes.

The 2015VictorianGender Equality Strategy alignswith our recommendations for

policyoutlinedabove.Specifically,thissetsoutsixsettingsforstate-wideactionon

genderequality,includingineducationandtraining,work,andleadershiproles

(VictorianGovernment2015).

In stating our endorsement of this strategy however, we would also offer some

caveats. The results of our review suggest that the relationship between gender

equality and health is not linear, nor does it guarantee universally better health

outcomes.While,overall,webelievethebulkoftheevidenceisinfavourofgender

equality, therewere a number of circumstanceswhere theremay be increases in

47

poor health behaviours such as smoking or alcohol use. To some extent, these

results may reflect the differences in progression of gender equality in different

spheresof lifeandover time.Forexample,an increase in femaleparliamentarians

doesnotnecessarilymeanthattraditionalgenderrolesinthehome(wherefemales

takeon thebulkofhomeandchildwork)willbe immediatelyaltered in favourof

more equal responsibilities between males and females. Further, gender equality

implementedatapolicyleveldoesnotnecessarilyflowtoinfluenceculturalnorms

abouthowwomenandmenoperate in a varietyof life arenas. Inour review, the

studiesthatmeasuredgenderequalityinanumberofdifferentdomains,includingin

thehome, intheworkplace,and inpoliticsprovidethemostcompletemeasureof

gender equality. We would suggest that similar measurement approaches are

undertakeninAustralianresearchtoassessgenderequalityanditseffectsonhealth.

On this point, we would argue that action for gender equality needs to occur at

multiple levels in society, including at work, in the home, and in the community.

Policy alone is not enough. Furthermore, any interventionsmust acknowledge the

broadersocialcontextinwhichtheyaretobedelivered,aswellasthewaysthatthe

social contextcan interactwithgenderequality tosometimesproduceparadoxical

results.

Changestowardsgenderequalitywilltaketime,however,implementingsupportive

policy for gender equality is a critical building block from which societal and

individual change can occur. As a final recommendation, we would suggest the

need for policy orientated research to understand the health impacts of gender

equalityontheVictorianpopulationovertime.Itisworthwhileconsideringthatthe

48

experienceandmeaningofgenderequality(anditsrelationshipwithhealth)islikely

to vary depending on a range of socio-economic, cultural and geographic factors.

Thus,wesuggestthatthereneedstobearangeoffurtherstudiesconductingaimed

atassessinggenderequalityandhealthacrossa rangeofdifferentcontextswithin

Australia.

49

References

Aizer, A. (2010). "The Gender Wage Gap and Domestic Violence." Am Econ Rev100(4):1847-1859.Backhans,M.,B.Burström,A.P.deLeonandS.Marklund(2012)."Isgenderpolicyrelated to the gender gap in external cause and circulatory disease mortality? Amixed effectsmodel of 22 OECD countries 1973–2008." BMC Public Health12(1):969.Backhans,M.C.,B.Burström,L.LindholmandA.Månsdotter(2009)."Pioneersandlaggards- Istheeffectofgenderequalityonhealthdependentoncontext?"SocialScienceandMedicine68(8):1388-1395.Backhans,M. C.,M. Lundberg and A.Mansdotter (2007). "Does increased genderequalityleadtoaconvergenceofhealthoutcomesformenandwomen?AstudyofSwedishmunicipalities."SocialScience&Medicine64(9):1892-1903.Balish,S.M.,R.O.Deaner,S.Rathwell,D.RainhamandC.Blanchard(2016)."Genderequality predicts leisure-time physical activity: Benefits for both sexes across 34countries."CogentPsychology3(1).Bentley,R.andA.M.Kavanagh(2008)."Genderequityandwomen'scontraceptionuse."AustralianJournalofSocialIssues43(1):65-80.Bilal,U.,P.Beltran,E.Fernandez,A.Navas-Acien,F.BolumarandM.Franco(2016)."Gender equality and smoking: a theory-driven approach to smoking genderdifferencesinSpain."TobControl25(3):295-300.Bond,J.C.,S.C.Roberts,T.K.Greenfield,R.Korcha,Y.YeandM.B.Nayak(2010)."Gender differences in public andprivate drinking contexts: amulti-levelGENACISanalysis."IntJEnvironResPublicHealth7(5):2136-2160.Borrell,C.,L.Palència,C.Muntaner,M.Urquía,D.MalmusiandP.O'Campo(2014)."Influence of macrosocial policies on women's health and gender inequalities inhealth."EpidemiologicReviews36(1):31-48.Bosque-Prous,M.,A.Espelt,C.Borrell,M.Bartroli,A.M.Guitart,J.R.VillalbíandM.T.Brugal (2015). "Genderdifferences inhazardousdrinking amongmiddle-aged inEurope:theroleofsocialcontextandwomen'sempowerment."EuropeanJournalofPublicHealth25(4):698-705.Brinda,E.M.,A.P.RajkumarandU.Enemark(2015)."Associationbetweengenderinequality indexandchildmortality rates:across-nationalstudyof138countries."BMCPublicHealth15(97):(9February2015).Chandola, T.,H. Kuper, A. Singh-Manoux,M.Bartley andM.Marmot (2004). "Theeffect of control at home on CHD events in the Whitehall II study: GenderdifferencesinpsychosocialdomesticpathwaystosocialinequalitiesinCHD."SocSciMed58(8):1501-1509.Chen,Y.Y.,S.V.Subramanian,D.Acevedo-GarciaandI.Kawachi(2005)."Women'sstatus anddepressive symptoms: amultilevel analysis." Social Science&Medicine60(1):49-60.Chon,D.S.(2016)."Aspuriousrelationshipofgenderequalitywithfemalehomicidevictimization:Across-nationalanalysis."Crime&Delinquency62(3):397-419.Clark, R. andB.M. Peck (2012). "Examining theGenderGap in Life Expectancy: ACross-NationalAnalysis,1980-2005."SocialScienceQuarterly93(3):820-837.

50

CSDH (2008).Closing thegap inageneration:Healthequity throughactionon thesocialdeterminantsofhealth.Finalreportofthecommissiononsocialdeterminantsofhealth.Geneva,WorldHealthOrganization.Dahlin, J. and J.Harkonen (2013). "Cross-national differences in the gender gap insubjective health in Europe: Does country-level gender equality matter?" SocialScience&Medicine98:24-28.Dereuddre, R., S. Van de Velde and P. Bracke (2016). "Gender inequality and the'East-West'divideincontraception:Ananalysisattheindividual,thecouple,andthecountrylevel."SocialScience&Medicine161:1-12.DiezRoux,A.V.(2002)."Aglossaryformultilevelanalysis."JournalofEpidemiologyandCommunityHealth56(8):588(587).Eek,F.andA.Axmon(2015)."Genderinequalityathomeisassociatedwithpoorerhealthforwomen."ScandJPublicHealth43(2):176-182.Elwér, S., L. Harryson,M. Bolin and A. Hammarström (2013). "Patterns of GenderEqualityatWorkplacesandPsychologicalDistress."PLoSONE8(1).Goode,W.J.(1960)."Atheoryofrolestrain."AmericanSociologicalReview25:483–496.Gressard,L.A.,M.H.SwahnandA.T.Tharp(2015)."Afirstlookatgenderinequalityas a societal risk factor for dating violence." American Journal of PreventiveMedicine49(3):448-457.Grittner, U., S. Kuntsche, K. Graham and K. Bloomfield (2012). "Social inequalitiesandgenderdifferencesintheexperienceofalcohol-relatedproblems."AlcoholandAlcoholism47(5):597-605.Hammarström,A.andS.P.Phillips(2012)."Genderinequityneedstoberegardedasa social determinant of depressive symptoms: Results from theNorthern Swedishcohort."ScandinavianJournalofSocialMedicine40(8):746-752.Harryson, L., M. Novo and A. Hammarstrom (2012). "Is gender inequality in thedomestic sphere associated with psychological distress among women and men?Results from theNorthern Swedish Cohort." J Epidemiol CommunityHealth66(3):271-276.Hassanzadeh, J., N. Moradi, N. Esmailnasab, S. Rezaeian, P. Bagheri and V.Armanmehr (2014). "The correlationbetween gender inequalities and their healthrelated factors in world countries: a global cross-sectional study." EpidemiologyResearchInternational2014.Johansson, K., P. Wennberg and A. Hammarström (2014). "Parental leave andincreasedphysicalactivityoffathersandmothers-resultsfromtheNorthernSwedishCohort."EuropeanJournalofPublicHealth24(6):934-939.Jun,H.-J.,S.Subramanian,S.GortmakerandI.Kawachi(2003)."Amultilevelanalysisofwomen's statusandself-ratedhealth in theUnitedStates." JAmMedWomensAssoc(1972)59(3):172-180.Kawachi,I.,B.P.Kennedy,V.GuptaandD.Prothrow-Stith(1999)."Women'sstatusandthehealthofwomenandmen:aviewfromtheStates."SocSciMed48(1):21-32.Kim, J. I.andG.Kim(2014). "Labor forceparticipationandsecondaryeducationofgenderinequalityindex(GII)associatedwithhealthylifeexpectancy(HLE)atbirth."IntJEquityHealth13:106.

51

Korpi,W.(2010)."Classandgenderinequalitiesindifferenttypesofwelfarestates:The Social Citizenship Indicator Program (SCIP)." International Journal of SocialWelfare19(SUPPL.1):S14-S24.Lahelma,E.,S.Arber,K.KiveläandE.Roos(2002)."MultiplerolesandhealthamongBritish and Finnishwomen: the influence of socioeconomic circumstances." SocialScience&Medicine54(5):727-740.Liberati,A.,D.G.Altman,J.Tetzlaff,C.Mulrow,P.C.Gøtzsche,J.P.A.Ioannidis,M.Clarke,P.J.Devereaux,J.KleijnenandD.Moher(2009)."ThePRISMAstatementforreportingsystematicreviewsandmeta-analysesofstudiesthatevaluatehealthcareinterventions:explanationandelaboration."BMJ339.Macintyre, S. and A. Ellaway (2003). Neighbourhoods and Health: An Overview.Neighbourhoods and Health. I. Kawachi and L. F. Berkman. New York, OxfordUniversityPress:20-42.Månsdotter, A., L. Lindholm, M. Lundberg, A. Winkvist and A. Öhman (2006)."Parental share in public and domestic spheres: A population study on genderequality,death,andsickness."JournalofEpidemiologyandCommunityHealth60(7):616-620.Månsdotter, A.,M. Lundberg and L. Lindholm (2012). "How does gender equalityprogress link to alcohol care and death? A registry study of the Swedish parentalcohortof1988/1989."JournalofPublicHealthPolicy33(1):105-118.Månsdotter, A.,M.Nordenmark andA.Hammarström (2012). "The importanceofchildhood and adulthood aspects of gendered life for adult mental ill-healthsymptoms – a 27-year follow-up of the Northern Swedish Cohort." BMC PublicHealth12(1):493.Mark, Q. J. (2014). "Global variance in female population height: the influence ofeducation, income, human development, life expectancy, mortality and genderinequalityin96nations."JournalofBiosocialScience46(1):107-121.Mayer,P. (2000). "Development,genderequality,andsuicide rates."PsychologicalReports87(2):367-372.McLaughlin,K.A.,Z.Xuan,S.V.SubramanianandK.C.Koenen(2011)."State-levelwomen'sstatusandpsychiatricdisordersamongUSwomen."SocialPsychiatryandPsychiatricEpidemiology46(11):1161-1171.Muntaner,C.,C.Borrell,A.Espelt,M.Rodríguez-Sanz,M.I.PasarÍn,J.BenachandV.Navarro (2010). "Politics or policies vs politics and policies: a comment onLundberg."Internationaljournalofepidemiology39(5):1396-1397.Neyer,G.,T.LappegǻrdandD.Vignoli (2013)."Genderequalityandfertility:whichequalitymatters?"EuropeanJournalofPopulation29(3):245-272.Norström, L., L. Lindberg and A. Månsdotter (2012). "Could gender equality inparental leave harm off-springs' mental health? a registry study of the Swedishparental/childcohortof1988/89."InternationalJournalforEquityinHealth11(1).Palència,L.,D.DeMoortel,L.Artazcoz,M.Salvador-Piedrafita,V.Puig-Barrachina,E.Hagqvist,G. Pérez,M. E.Ruiz, S. Trujillo-Alemán,C.Vanroelen,D.Malmusi andC.Borrell(2017)."GenderPoliciesandGenderInequalitiesinHealthinEurope:ResultsoftheSOPHIEProject."InternationalJournalofHealthServices47(1):61-82.Pampel, F. C. (2001). "Gender equality and the sex differential in mortality fromaccidents in high income nations." Population Research and Policy Review 20(5):397-421.

52

Platt,J.M.,S.J.Prins,L.M.BatesandK.M.Keyes(2016)."Unequaldepressionforequalwork?How thewage gap explains gendered disparities inmood disorders."SocialScience&Medicine149:1-8.Redding,E.M.,M.T.Ruiz-Cantero,J.Fernandez-SaezandM.Guijarro-Garvi(2016)."Gender inequality and violence against women in Spain, 2006-2014: towards acivilizedsociety."GacSanit26(16):30192-30193.Ricketts, C. F. (2014). "Re-Examining the Gender Gap in Life Expectancy: a CrossCountryAnalysis."InternationalJournalofHumanitiesandSocialScience4(10):38-51.Roberts, S.C.M. (2012). "Macro-level genderequalityandalcohol consumption:amulti-levelanalysisacrossU.S.States."SocialScience&Medicine75(1):60-68.Sainsbury,D.(1999).Genderandwelfarestateregimes,OxfordUniversityPress.Sanz-Barbero, B., C. Vives-Cases, L. Otero-García, C. Muntaner, J. Torrubiano-Domínguez and P. O'Campo (2015). "Intimate partner violence among women inSpain: the impact of regional-level male unemployment and income inequality."EuropeanJournalofPublicHealth25(6):1105-1111.Saunders,M.H.andA.Peerson(2013)."Genderequityinhealthpolicy:responsetoKeleher."AustNZJPublicHealth37(5):491-492.Sen,G.andP.Östlin (2008). "Gender inequity inhealth:why itexistsandhowwecanchangeit."GlobalPublicHealth3(sup1):1-12.Shah, A. (2008). "Association of suicide rates for elderly age bands with genderequality."PsycholRep102(3):887-892.Sörlin,A.,A.Hman,N.NgandL.Lindholm(2012)."Cantheimpactofgenderequalityon health be measured? a cross-sectional study comparing measures based onregisterdatawithindividualsurvey-baseddata."BMCPublicHealth12(1).Sörlin,A.,L.Lindholm,N.NgandA.Öhman(2011)."Genderequalityincouplesandself-ratedhealth-Asurveystudyevaluatingmeasurementsofgenderequalityanditsimpactonhealth."InternationalJournalforEquityinHealth10:37-37.Sorlin, A., A. Ohman and L. Lindholm (2011). "Sickness absence in gender-equalcompaniesaregisterstudyatorganizational level."BMCPublicHealth11(548):(11July2011).Tesch-Romer,C.,A.Motel-KlingebielandM.J.Tomasik(2008)."Genderdifferencesinsubjectivewell-being:Comparingsocietieswithrespecttogenderequality."SocialIndicatorsResearch85(2):329-349.Thoits, P. A. (1983). "Multiple identities and psychological wellbeing." AmericanSociologicalReview48:174-187.UnitedNationsDevelopmentProgramme(2010).HumanDevelopmentReport2010.NewYork,USA,UnitedNations.Van Tuyckom, C., S. Van de Velde and P. Bracke (2012). "Does country-contextmatter? A cross-national analysis of gender and leisure time physical inactivity inEurope."TheEuropeanJournalofPublicHealth:cks009.Velde, S. v. d., T. Huijts, P. Bracke and C. Bambra (2013). "Macro-level genderequality and depression in men and women in Europe." Sociology of Health andIllness35(5):682-698.VicHealth(2013).VicHealthActionAgenda.Melbourne,VictorianHealthPromotionFoundation.

53

VictorianGovernment(2015)."SAFEANDSTRONGAVICTORIANGENDEREQUALITYSTRATEGY."http://www.vic.gov.au/system/user_files/Documents/women/161108_Victorian_Gender_Equality_Strategy_ONLINE.pdf.Wängnerud,L.andA.Sundell(2012)."Dopoliticsmatter?WomeninSwedishlocalelectedassemblies1970–2010andgenderequalityinoutcomes."EuropeanPoliticalScienceReview4(1):97-120.Wells, J. C. K., A. A. Marphatia, T. J. Cole and D. McCoy (2012). "Associations ofeconomicandgender inequalitywithglobalobesityprevalence:understanding thefemaleexcess."SocialScience&Medicine75(3):482-490.WHO (2015). "Media Centre: Gender. Fact sheet N°403."http://www.who.int/mediacentre/factsheets/fs403/en/;.Witvliet,M.I.,O.A.Arah,K.StronksandA.E.Kunst(2014)."Aglobalstudyonlonemothers:Exploringtheassociationsofself-assessedgeneralhealthwithmotherhoodtypes and gender inequality in 32 countries."Women's Health Issues24(2): e177-e185.World Economic Forum (2014). The Global Gender Gap 2014. Switzerland,WorldEconomicForum.World Economic Forum (2017). "The Global Gender Gap Report 2016."http://reports.weforum.org/global-gender-gap-report-2016/.Yllö, K. (1983). "Sexual equality and violence against wives in American states."JournalofComparativeFamilyStudies:67-86.Yllö,K.(1984)."Thestatusofwomen,maritalequality,andviolenceagainstwives:Acontextualanalysis."JournalofFamilyIssues5(3):307-320.

54

Table2.DetaileddescriptionofthestudiesincludedinthereviewAuthor Population/setting Participants Objectives Studydesign Outcome Exposures: Gender

equalityindicatorsConfounders/covariates Mainresults

WorkandemploymentrelatedgenderequalityAizer,2010 Population and

administrativedata:Administrative data onfemale hospitalisationfor assault inCalifornia.County level data onwages from Bureau ofEconomic Analysisannual survey ofemployers.

Workingpopulation ofCalifornia,n=?15millionwomen.

Examine associationsbetween gender wagegap and domesticviolence.

Longitudinal usingnon-linkedadministrative data(1990-2003)

Female hospitalisationduetoassault.

Genderwagegap. RaceUnemploymentrateNon-intimatehomicidesIncarcerationnumberStudentpopulationImmigrationrateGenderNonassaultinjuries

Reductioninthewagegapexplains9%ofreductioninviolenceagainstwomen.

Elweretal.,2013 Swedishcohortstartedfrom last year of highschool(1983),followedup with surveys andlinked to healthinsurance and labourmarket statisticsdatasets

NorthernSwedish Cohortn=715,Participantswereemployed malesandfemales

• Tounderstandthepatternsofgenderequalityandmentalhealth

• Genderequalityismultidimensional

Linked cohort design(1983to2007)

Psychological distress(menandwomen)

5 item scalerepresenting genderequality in aworkplace,representing ratio ofmale/female:• Numberofemployees

• Salary• Educationallevel• Parentalleave• Temporaryparentalleave

Previous psychologicaldistressSocio-economicpositionOccupationalclassificationProportionofemployeesyoungerthan38years

Patterns of genderinequality in workplaceswere related to women’sbut not men’spsychologicaldistress.In workplaces that wereunequal and traditionalregarding gender,women’s mental healthsuffersmorethanmen’s.

Plattetal.,(2016)

US nationallyrepresentativesurvey

22,581 adults(30-65years)

To understand therelationshipbetweenthewage gap and mentalhealth

Cross-sectionalsurvey Past year and lifetimemajor depressive andanxiety disorders (menandwomen)

Individual incomebetween men andwomen (propensityscore matchedapproach)

Propensity scoreincluded: age, ageentered the workforce,educational attainment,whether born in theUS,occupation, full or parttime working, industry,employer type, maritalstatus, number ofchildrenathome

When female income waslower than those of amatchedmalecounterpart(adjusting for propensityscore), odds of depressionand anxiety elevatedamong women. Whenfemale income exceededor was the same asmatched counterpart, theodds of depression andanxiety was non-significant, but still

55

elevated.

Johansson et al.,2014

NorthSwedishCohort Parents in thecohort,n=584

To understandrelationship betweenparental leave andphysicalactivity

Linked cohort design(1983to2007)

Change in physicalactivity fromage 21 to42 (decreased, stable,increased) (men andwomen)

Parental leave inwomenandmen.

SES based onoccupation,yearofbirthofthechild

No relationship betweenlength of parental leaveandphysicalactivityatage42 for women. For men,longer parental leaverelated to greater physicalactivity.

Norstrom et al.,2012

PopulationofSweden Parents andchildren (50589boys and 48524girls)

To explore genderequality in childcare andmental ill-health amongchildren

Linked administrativedataset (SocialInsurance agencydataset, linked tooutpatient and drugregister)originalstudy

Outpatient services formentalhealthanddrugregister (anxiety anddepression) (men andwomen)

Parental leave –categorised into fivecategories- rangingfrom traditionallyunequal (mother 80%of the leave) to veryuntraditionallyunequal(father80%oftheleave)

Total number of days ofleave, year of birth,having another child,born outside Sweden,educational level,institutional care ofparents in a mentalhealthfacility.

No significant results.Compared to those whoshared parental leaveequally, higher odds ofdepression among girlswho had traditionalparental roles. Males hadlower odds of depressionand anxiety in all otherparental role groups (verytraditional, rathertraditional, untraditional)compared to parents thathadequalparentalleave.

Sorlinetal.,2011 PopulationofSweden 123 companies,53,204 workingmales andfemales

To understand whethergender equality reducesthe number of days ofsick leave at thecompanylevel

Linked administrativedataset, cross-sectionaldataset

Sickness absence (2 to4days–“daysoffsick”;15daysandover“dayson sickness benefit”)(menandwomen)

Organisational GenderGap Index (OGGI)-ratio at organisationallevel:• male/femaleemployeeratio

• percentageoffulltimeemployees

• educationallevel• monthlyincome• daysonparentalleave

• daysontemporalleave.

Age, education, income,full/parttime,andsector

Compared to genderunequal companies, therewere higher odds ofsickness absence amongthose employed inorganisations that wereequal. The differenceswere greater for menratherthanwomen.

Sorlinetal.,2012 PopulationofSweden Two sectors:section 1=11,471people in 46companies;sector 2=32,151

To understand theassociation betweengender equality atworkandself-ratedhealth.

Linked administrativedataset, cross-sectionaldataset

Self-reported health(menandwomen)

Organisational GenderGap Index (OGGI)-ratio at organisationallevel:• male/female

Age, education level,income, employmentlevel, type of companybased on the genderequalityindex

Women in gender equalcompanies reportedsignificantly better self-rated health than thoseworked in unequal

56

individuals in 77companies

• employeeratio,• percentageoffulltimeemployees,

• educationallevel,• monthlyincome,• daysonparentalleave,

• daysontemporalleave.

organisations. Noassociationsformen.

Withincountriesanalysis–genderequalitymeasuredwithinthefamilydomainNeyeretal.,2012 Multi-country

(Bulgaria, France,Germany, Georgia,Norway, Romania,Russia, Hungary, Italy,Netherlands)

First waves ofGenerations andGenderSurveys

To understand the linkbetween genderequality andchildbearingintentions

Cross-countrysurvey Child bearingintentions of womenandmen

• Employmentstatus,• precariousemployment,

• workingarrangement,

• Houseworkdivision,childcaredivision,

Age, education, maritalstatus, employmentstatus. Partners activitystatus,country

Women and men hadhigher childbearingintentions in more equalcouples

Mansdotter et al.,2012

Sweden All fathers andmothers whohad their firstchild in 1988 or1989, linked tonational healthdatasets.

How gender equalityduring (income andcaring roles) in earlyparenthood relates tosubsequent alcohol-related inpatient and/ormortality

Cohort, country levellinkageofdatasets.Gender equalitymeasured (1988-1991), alcohol relatedinpatient care ormortality(1992-2006)

Alcohol relatedinpatient care ormortality (men andwomen)

Three categories ofgenderequality:• Equality–bothparentshadbetween40%and60%ofindicatorsbelow

• Traditionalinequality(fatherhad>60%ofincomeandwomenhad>60%caringroles)

• Untraditionalinequality(mother’shad<40%caringandfathers<40%ofincome)

Dataincluded• Incomedata• Occupationalposition

• Parentalleavedata• Temporarychildcaredata

Age, birth outsideSweden, income 1990-1991, education 1990,married/cohabiting1990, other children,alcohol related care twoyears before and duringmeasurement of genderequality.

Overall, traditionalwomenrun lower risk, whereastraditional men anduntraditional women(those opposing thetraditional division ofparenthoodresponsibilities) runhigherrisk of alcohol harm thangender-equalcounterparts.

57

Mansdotter et al.,2006

Sweden All Swedishcouples (98240people)who hadfirstchildin1978

To understand therelationship betweengender equality andhealth

Cohort, country levellinkageofdatasets.

Overall mortality andsicknessabsence

Five categories ofgenderequality:• Pronouncedtraditional

• Moderatetraditional• Equal• Moderateuntraditional

• Pronounceduntraditional

IncomedataOccupationalpositionParentalleavedataTemporary childcaredata

Age, socioeconomicposition, income,countryofbirth

Equality in income andoccupationwas associatedwith increased odds ofmortalityforwomen.Noeffectsformen.For females, beingtraditional associated withlowersicknessrisks.

Backhans et al.,2009

279 Swedishmunicipalities

37,423 men and37,616womenin279 Swedishmunicipalities,who had theirfirst child in1978.

Does the contextimpacts on theassociation betweengender equality andhealth

Registered data – allSwedish couples(N=49120) first childin 1978, followed upuntiltheyear1999.

Compensated daysfrom sicknessinsurance during 1986-1999.

Gender equality atindividual/couplelevels:fivecategories:• Pronouncedtraditional-fathers>80%,mothers<20%

• Moderatelytraditional–father60-80%,mother20-40%

• Equal–eitherparent40-60%

• Moderatelytraditional–father20-40%,mother60-80%

• Pronounceduntraditional-father<20%,mother80&Gender equality atindividual/couplelevels:

• Politicalparticipation• Economicresources• Occupationalsexsegregation

Area level: totalemployment rate, andthe proportion of thepopulation withmandatory education.Individual level: type ofwork,Swedishorforeignborn.

Among men: those whoare equal in an equalmunicipality have lowerlevels of sick leave, whilethose who are less equalthan their municipality. Inthe public spheres, thosemore equal than themunicipalities fare betterthan average. Motherswhoaretraditionalintheirrolesinpublicsphereshadhigh levels of sick leave,reversetrueforthosewhoare equal. Traditionalmothers in traditionalmunicipalities have thelowest levels of sick leaveand pioneers had thehighest.

58

• Caringwork• Interactionbetweencoupleandarealevel

Mansdotter et al,2012

Sweden Participants inNorthernSwedish cohortstudy.At26yearfollow-up, n=481women and 526men)

Examine associationsbetween childhood andadulthood gendered lifeonmentalhealth

Cohortstudy Mental ill health(depressive symptoms,anxietysymptoms)

Gendered life:traditional or non-traditional inchildhood (parentsgendered life),adulthood at age 30(gender ideology andattitude to childcare),adulthood age 42partnership equalityand attitude tochildcare)

Self-reported anxietyand depression at age16, perceived childhoodclassatage16,divorcedor dead parents at age16, educational level atage 30,married/cohabiting atage 30, number ofchildrenatage30.

1. Genderequalityinadulthoodassociatedwithgoodself-reportedmentalhealthregardlessofchildhood(parentsgenderedlife)

Reduced depressivesymptoms at age 42 inmenwho regard childcareasfemalerole

59

Chandola et al,2004

UnitedKingdom Data fromphases 3-5 ofWhitehall IIstudy,n=7470

• Investigatethemeaningofcontrolathome,

• ExaminetheeffectofcontrolathomeonincidentCHDevents

Ascertain the extent towhich control at homeexplains socialinequalities in CHDevents

Cohort study of civilservants

FatalCHDNon-fatal myocardialinfarction

Controlathome Civilservicegrade,Household financialproblemsDemandoverloadPowerwithinhousehold

1. LowcontrolatcomepredictsCHDamongwomenbutnotmen

Sorlinetal,2011 Sweden 1400participantsdrawnfrompriorregistersurveyof1.1 millionpeople workingin 8000companies inSweden

Analyse associationbetween genderequality in partnerrelationship and self-ratedhealth

Cross-sectional SRH Self reported genderequality indexmeasuredacrossthreedomains(1.education,income. full/part timeemployment; 2.Sharing ofresponsibilities andtime; 3. Sharing ofparental leavefollowingbirthofchildand temporary leavefor sickness of child)for respondent andpartnerSelf-perceived genderequalityindex

AgeEducationOccupationalgradeIncome

1. Nosignificantassociationbetweenself-reportedgenderequalityindexandSRH

2. Formen,significantassociationbetweenperceptionofgenderequalityinpartnerrelationshipandgoodSRH

3. Forwomenthereweresimilarassociationsbutthesewerenon-significant

4. Menperceivedgreatergenderequalitythantheyreported

Women perceived lessequality than theyreported

60

Eek and Axmon,2015

Sweden 837 womenliving inrelationship andworking at least50%offulltime.Participantsweredrawn from anoriginal surveyamong parentswhose youngestchildwas2yearsold

Investigate whetherunequal distribution ofresponsibilities in homeis related to healthoutcomesinwomen

Cross-sectional Subjectiveglobal stress(PSS)Physical/psychosomaticsymptoms(13commonhealthcomplaints)General physical andmental self-ratedhealthWorkrelatedfatigueSatisfaction withgeneral life and worksituationsWorkstressWorkengagementWorktofamilyconflictFamilytoworkconflictLeisuretimeactivities

Division of householdduties

Employment status(full/parttime)Educationallevel

Women in relationshipswho perceived greaterinequalityindistributionofhousehold responsibilitieshad significantly higherself-perceived stress,fatigue, physical andpsycho-somatic symptomsand work family conflictscomparedtowomenlivingin more equalrelationships.

Hammerstrom etal,2012

Sweden Participants inNorthernSwedish cohortstudy who werecohabiting ormarried at 26year follow-up,n=390 womenand383men)

Analyse importance ofgenderrelationsfor

Cohortstudy Depressive symptomsatage42

Perceptions of genderinequity in partnerrelationshipHouseholdresponsibilitiesSocioeconomicstatusFinancialstrain

Prior depressivesymptoms

1. Depressivesymptomsweresignificantlyassociatedwithfinancialstrain

Amongwomen,depressivesymptoms weresignificantly associatedwith perceived genderinequity

61

Harryson et al,2012

Sweden Participants inNorthernSwedish cohortstudy who werecohabiting at 26year follow-up,n=372 womenand352men)

Examine associationsbetween genderinequality in householdis associated withpsychological distressformenandwomen

Cohortstudy Psychologicaldistress Gender inequalitywithin household(measured asperceptions of genderinequality, time spenton household chores,time spent onhousehold work,responsibility fordomestic work andchildcare).

Psychological distress atage21TimeinpaidworkSocioeconomicpositionNumberofchildren

1. Perceivedgenderinequalityincouplerelationshipwassignificantlyassociatedwithpsychologicaldistressformenandwomen

2. Forwomen,beingsolelyresponsiblefordomesticworkwassignificantlyassociatedwithgreaterpsychologicaldistress

3. Formen,havingresponsibilityforlessthanhalfdomesticworkwassignificantlyassociatedwithgreaterpsychologicaldistress

62

Withincountries-genderequalityinpolitics,theeconomy,socialservices,andreproductiverightsSanz-Barberoetal.,2015

Spain 7898 adultwomen

To understand the roleof contextual levelunemployment andgender based incomeinequalityonIPV

Cross-countrysurvey IPV(women) Gender basedeconomicdiscrimination (GDPpercapitaminusindexof income separatelyforwomenandmen)

Individual level variables(age, citizenship, socialclass, education,mothers exposure toIPV) and regional levelvariables (male longterm unemployment,Gini coefficient ofincomeinequality)

Women residing In areaswith high gender-basedincome discrimination hadlower IPV compared tothose in areas with lowgender-based incomediscrimination.

Wangnerud et al.,2012

Swedish local electedassemblies

238municipalities

Tounderstandtheeffectof female councillors(1985-2010) on poorhealth

Longitudinal pooledecological analysis,1970-2010

Poorhealth(women) • Femalecouncillors(%)

• Left-greenparties(%)

• Femalevoterturnout

• Publicemployee• Femaleeducation

Average municipalpopulation, geographicalarea in squarekilometres, averagepercentage of womenand men with tertiaryeducation.

In municipalities wherefemaleturnoutwashighinrelationtowomen,womenhadbetterhealth.

McLaughlin et al.,2011

UnitedStates Nationalprobabilitysample of USadults(n=34,653), datafrom theNationalEpidemiologicalSurvey onAlcohol andRelatedConditions

To what extent is statelevel women’s statusrelated to psychiatricdisorders inwomenandgender differences inpsychopathology

Cross-sectionalsurvey 12-month mood andanxiety disorders(women)

Women’sstatusatthestatelevels:• Politicalparticipation• Employmentearnings

• Social/economicautonomy

• Reproductiverights

Age, race/ethnicity,marital status,educational attainment,household income,occupation.

The prevalence of majordepression and PTSD waslower in states wherewomen have reproductiverights. Other variables ofstate level women’s rightvariables were unrelatedtodepressionandanxiety.

Robertsetal.,2012 UnitedStates Surveydatafromthe 2005Behavioural RiskSurveillanceSystem

To understand therelationship betweenstate-level genderequality and alcoholconsumption

Cross-sectionalsurvey Alcohol consumptionincluding past 30 daydrinker status, drinkingfrequency, bingedrinking, volume, riskydrinking (men andwomen)

Statelevel• gendersocioeconomicstatus

• reproductiverights• policiesrelatingtoviolenceagainstwomen

women’s politicalparticipation

State level controls-income inequality,religion. Individualcontrols – age, race,income, marital status,education, employmentstatus.

All gender equalityindicators were positivelyassociated with women’sand men’s drinker status.Findings do not supportthe hypothesis that highgender equality onwomen’s status isassociated with higheralcohol consumption. Infact, higher genderequality was associated

63

with less alcoholconsumption.

Chenetal.,2005 FiftyAmericanstates 7789womenina1991longitudinalfollow up study(participantswereanationallyrepresentativerandom sampleof women whogavebirth to livebabiesin1988)

To assess therelationship betweenstate level women’sstatus variables andindividual depressivesymptoms

Follow up study ofparticipants of anationallyrepresentativesample

Symptoms ofdepression (CES-D)(women)

Composite indicatorsmeasuring:• Women’spoliticalparticipation

• Economicautonomy• Employmentandearningsand

Reproductiverights.

Age,race,familyincome,educational attainment,employment status,marital status, incomedistribution at the statelevel,GINIcoefficient.

Women residing in stateswith high scoresemployment and earningon index had lowerdepressionthanthosewhoscoredlowerontheindex.Women in states withhigherscoresoneconomicautonomy had lowerdepression. Women instates with highreproductive rights hadlowerdepression.

Backhans et al,2007

Population statisticsfrom 289 Swedishmunicipalities

n=289 Swedishmunicipalities

To test the hypothesisthat greater genderequality is associatedwith better healthoutcomes

Cross-sectional studyofpopulationdata

LifeexpectancyNumber compensateddays per insuredperson for sicknessabsence and disability(menandwomen)

• Politicalparticipation:proportionofwomenvsmeninmunicipalcouncilsandexecutivecommittees

• Temporaryparentalleave

• Proportionofparttimeworkers

• Proportionofmenandwomeninhealthcareandsocialservices

• Proportionofmenandwomeninmanufacturing

• Proportionofmenandwomeninmanagerialpositions

• AverageincomeRelativepoverty

MunicipalityfinancesLocallabourmarketLocaltaxrateRobinHoodtaxTotalemploymentrateProportionofpopulationwith any post-secondaryeducationAgestructureProportionstudents

Genderequalityassociatedwith higher levels ofsickness and disability formenandwomen.Genderequalityassociatedwithlowerlifeexpectancy.

Kawachietal,1999 USA Age-standardisedcause-specificmortality from

Examine associationsbetween the status ofwomenandwomenand

Ecological and crosssectional

Total female and malemortality ratesFemale cause specific

State level status ofwomen assessed byfourcompositeindices

Statespecific:PovertyratesIncomedistribution

Higher politicalparticipation by womencorrelated with lower

64

50USstates men’shealthstatus deathratesMean days of activitylimitations reported bywomen in thepreviousmonth

thatmeasured:• politicalparticipation• economicautonomy• employmentandearnings

• reproductiverights

GINIcoefficientsHouseholdsizeMedian householdincome

femalemortality ratesandloweractivitylimitationsSmaller wage gapassociated with lowerfemalemortality ratesandloweractivitylimitationsIndices of women’s statuswerecorrelatedwithmalemortalityratesAssociations betweenindices of women’s statusand femalemortality ratespersisted after adjustmentfor income inequality,poverty rates and medianhouseholdincome.

Junetal,2004 USA Women inBRFSSstudyinUSAN=87848

Examine associationsbetween self-ratedhealth and women’sstatus

Ecological and crosssectional

Self-ratedhealth State level status ofwomen assessed byfourcompositeindicesthatmeasured:• politicalparticipation• economicautonomy• employmentandearnings

reproductiverights

Statespecific:GINIcoefficientsMedian householdincome

Low status for women (asmeasured on politicalparticipation,employment, economicautonomy)was associatedwith higher likelihood ofreportingpoorhealth.

Yllo,1983 USA Individualscohabiting in 30USstates

Investigate associationbetween women’sstatusandIPV

Ecological and crosssectional

Proportion of coupleswho reported thathusband had usedviolenceagainstwife

Dimensions of Statusof Women’s index:economic, educationalandpolitical,legal.

Borderline significance fornegative relationshipbetween high legal statusforwomenandIPV.

Yllo,1984 USA Individualscohabiting in 30USstates

Investigate whethersexual inequality isassociated with therelationship betweenmarital inequality andIPV

Crosssectional Proportion of coupleswho reported thathusband had usedviolenceagainstwife

Dimensions of Statusof Women’s index:economic, educationalandpolitical,legal.

Higher levels of IPV incouples where husbanddominatesdecisionmakingresiding in states in whichthere is high status forwomen.Alsohigher levelsof IPVincouples with dominantwife, residing in stateswithlowstatusforwomen(lowgenderequity)

65

66

Betweencountries,fertilityrates,educationalstatusofwomen,labourforcestatusClarketa.,2012 Cross-countrystudy (1) Europe and

theWest (2) LatinAmerica andCaribbean (3)Central and Sub-Saharan (4) NorthAfrica and theMiddle East (5)East Asia and thePacific (6) EasternEurope andCentral Asia. 139countries

Predictorsofthegendergap in life expectancyacross large crossnationalsample

Cross-country panelstudy (Time 1985,1990, 1995, 2000,2005)

Gender differences inlifeexpectancyGender ratio in lifeexpectancy

Women’sstatus:• Fertilityrate• Labourforceratio-female

• Parliamentratio-female

GDPpercapitaGinicoefficientYearWorldregionWorld DevelopmentIndicator

As females begin toparticipate in traditionalmale institutions, such asformal schooling and paidemployment, andbegin toshiftawayfromtraditionalfemale roles, as indicatedbyadeclineinfertility,thisincreaseinwomen’sstatusuniformly serves toincrease female mortality.There needs to be greaterstudies into the long termmortalityeffectsofgenderequity.

SooChon2013 Cross-countrycountry 124countries To examine the role ofwomen’s educationalandrelativesocialstatuson the female homiciderate

Cross-country panelstudy,2002and2004

• Femalehomicidevictimizationrate

• Female-to-overallhomicidevictimisationratio.

Absolute genderequality-index• Femaleeducationalstatus

• Femaleincome• Femaleeconomicactivity

Relative genderequality• Ratiooffemale’scombinedgrossenrolmenttothatofmales

• Femalepoliticalempowerment-%offemalesinnationalparliament

• Femaleeconomicactivityasa%ofthemaleactivity.

GDPpercapitaGini coefficient ofincomeinequality.Ethnic heterogeneityindex% of the femalepopulation among thetotalnationalpopulation%offemalepopulation%ofthepopulationaged20to34yearsHuman DevelopmentIndexLow, medium and highHuman DevelopmentIndex

Non-significantrelationship betweengender equality andfemale homicidevictimisation rate.Criticises previous studiesand finds that therelationship between bothabsolute and relativegender equitywith femalehomicide rate may bespurious as they did notcontrol importantconfounders.

Bentleyetal.,2008 Within countryanalysis of postcodedistricts (UnitedKingdom) and Cross-

18 countries thatparticipatedintheFertility andFamily surveys

To examine whetherwomen’s labour forceparticipation andpolitical representation

Within country studyof postcode districts,and betweencountries analysis.

Contraception use intheNSSALContraceptionuseIn the Fertility and

Individual variables intheNSSAL:Age, marital status,number of children,

Age, relationship status,indicatorofemploymentstatus.

Gender equity may be animportant determinant ofcontraceptive use. In theFFS contraceptive usewas

67

countrystudy conducted by theUnited Nations inthe1990sAlsoananalysisofpostcode (646)areas in theUnited Kingdom(National Surveyof SexualAttitudes andLifestyles-NSSAL)

in parliament wasassociated withcontraceptionuse.

Both cross-sectionalpanels.

FamilyService(FFS) ethnicity, religion,attendanceatreligiousservices, education,employment statusand household socialclass, indicators ofsexualactivity, fertilitystatus, and desire forchildren.Area level variables intheNSSAL:percentageof women who wereemployed on a parttimeorfulltimebasis.

lower in countries wherelabour force participationwashighest.

Ricketts2014 Cross-countrystudies WorldDevelopmentIndicators, 99countries,1995to2010

To understand theeffects of increasedfemale labour forceparticipation on the lifeexpectancygendergap

Between countriesstudies (n=99), 1995to2010

Ratio: Life expectancyforfemales/Life expectancy formales

Labour forceparticipation (%of thefemale populationages 15 and older isthat is economicallyactive)

Age, population density,rural population growth,urban populationgrowth, urbanpopulation %, lowincome, lower middle,upper middle, highincome non-OECD, highincome OECD, year,education, foodproduction, consumerpriceindex,GDPgrowth,GDP per capita, capitalformation,inflation

Female labour forceparticipation rate issignificantly related to thegapinlifeexpectancy(e.g.,a 1% increase in femalelabour force participantsrate is association with adecrease in the lifeexpectancy gender gap of0.03%). This suggests isthat the labour forceparticipation rates ofwomen approach those ofmen, women’s lifeexpectancy is alsoexpected to approachthose of men, and viceversa.

Pampel2001 Highincomecountries18 high incomecountries,1955to1994

To investigate the“reduction inprotection” argumentpredicts that the femaleadvantage in accidentmortality will reduceover time, and theindicators of gender

Cross-country panelstudy,1955to1994

Numberofdeathsfromaccidents and suicideper100,000populationofspecifiedageandsexgroup

Female status in legal,family, political,occupation andeducationstatus.

Time, divorce, marriage,fertility

The reversal in thedownward trends in thefemale advantage in bothsuicide mortality - favourinstitutional adjustmenthypothesis (femaleadvantage will firstdecline,but then increase,

68

equality will reduce thefemaleadvantage.

and that indicators willreduce the femaleadvantage at low levels ofgender equality, butincrease the femaleadvantage at higherlevels).

Tesch-Romer et al,2008

57countries(WVS)5countries(OASIS)

Two comparativedatasets: WorldValues Survey &OASISproject(25 years andolder living inprivatehouseholds)

Examine whethersocietal genderinequality is related togender differences insubjectivewellbeing

Subjectivehealthrating(WVS)

Genderequalitynorms(WVS)Genderempowermentmeasure(GEM)

General lifesatisfaction(WVS)Welfareregimetype

Relationship betweensocietal gender equalityand SWB varies accordingto countries culture ofgender equality: if themajority of society rejectgender inequality on thelabour market, genderdifferences in SWBdecrease with increasingequality for both genderson the labour market; incountries where genderinequality on the labourmarket iswidelyaccepted,the gender differences inSWB increase with higherequality of women on thelabourmarket.

UNgenderequityindex:GIIBalishetal,2016

36countries 48154adultsfrom36 countries, allparticipants inISSP

• TounderstandwhethergenderequalityisrelatedtoLTPA,andexploreextenttowhichthisisgenderspecific.

Cross-sectionaldesign LTPA operationalisedas number of timesengaged in physicalactivity: 1/week ormore vs less than1/week (males andfemales)

Gender InequalityIndex(GII)

Individual: Age, level ofeducationCountry-level: GDP; %Muslim

Increased gender equalitysignificantly associatedwith higher LTPA for bothmaleandfemales.

Bilaletal,2015 Spain Spanish adultsaged 16+ yearsfrom nationallyrepresentativedatasources:National Instituteof Statistics (GII);national

• Examineassociationsbetweengenderinequalityandsmokingprevalencebygender,educationandbirthcohortovertheperiodof1960-2010

Ecological Smoking prevalence(maleandfemale)

Gender InequalityIndex(GII)

As gender inequalitydecreased over time, theratio of female to malesmokingincreased.

69

Parliamenthistorical sources(GII);National HealthInterview Survey(Smokingprevalence)

Brindaetal,2015 138 countries: 27 lowincome; 38 low-middle income; 30uppermiddleincome;43 high incomecountries

Populationstatisticsfrom138countries,obtained fromWHO statisticalinformationsystem

• Examineassociationsbetweengenderinequalityandchildmortality

Ecological • Infantmortality;• Neonatalmortality;• Under-5mortality;• Femaleunder-5mortality;

• Ruralunder-5mortality;

• Under-5mortalityduetoAIDS/HIV;

• Under-5mortalityduetodiarrhoea

Gender InequalityIndex(GII)

PercapitaGDPEconomic inequalityindex(Ginico-efficient);Immunisationcoverage

Greater gender inequalitysignificantly associatedwith higher neonatal,infant and under-5 childmortality rates. Theseassociationspersistedafteradjustingforeconomicandhealthservicevariables.

Dahlin&Harkonen,2013

28 Europeancountries

191104participants in 5waves (109country years) ofEuropean SocialSurveydata

• Assessrelationshipbetweengendergapinsubjectivehealthacross28Europeancountries

• Ascertainwhethergendergapsandcross-nationaldifferencesareexplainedbydemographicandsocioeconomicdifferences;

• Examinewhethernationallevelgenderequalityisrelatedtogendergapsinsubjectivehealth

Ecological SRH;LLImaleandfemale)

Gender;Gender InequalityIndex (GII) fitted asinteractionterm;

Individual: Age, age-squared, respondenteducation, partner’seducation, occupationalstatus, whetheroccupation missing,employment status,children, “tight withmoney”,calendaryearCountry level: incomeinequality (Gini); GenderInequality Index (GII);overall socialdevelopment(HDI)

1. Cross-nationalvariationingendergapsinhealth

2. CountrieswithgreatestgendergapinhealthareinEasterandSouthernEurope(worseforwomen);inothercountriesthereislittleornodifference

3. Genderinequalitynotassociatedwithgendergapsinhealth

4. HigherlevelsofhumandevelopmentwasassociatedwithreducedgendergapsinSRH(butnotLLI)

Dereuddre et al,2016

Data from 17Europeancountries

N=31632 womenaged 18-49 whohad a malepartner at the

• Assesswhethergenderinequalityisassociatedwithgreateruseofnoor

Cross-sectionalecological

Contraception type(female)

Women’s individualsocioeconomicposition (educationattainment and

Individual: Age, age-squared, partner status;parity;Urbanity (proxy for

1. Positiveassociationbetweenwomen’sindividualSEPanduseofmodernreversible

70

time. DatadrawnfromtheGGSandDHSsurveys.

traditionaldifferencesincontraceptiveuseareassociatedwithdifferencesinwomen’sstatusatindividuallevel,andgenderinequalityatcoupleandcountrylevel

employmentstatus);Women’s relativesocioeconomicposition (compared topartner);Gender InequalityIndex(GII)

supply of moderncontraceptives)Country level: incomeinequality(Gini)

contraceptives2. Negativeassociationbetweenwomen’sindividualSEPandfemalesterilization.

3. Equalityofoccupationrelativetopartnerassociatedwithhigheruseofmodernreversiblecontraception

4. Greatercountrylevelgenderequalityassociatedwithgreaterlikelihoodofusingmodernreversiblefemalemethods,butnotsterilisation

5. CentralandEasternEuropeancountrieshavehigherprevalenceofnocontraceptionortraditionalandmodernreversiblemalecontraceptionappearedtobeexplainedbycountrylevelgenderequalityindices.

Reddingetal,2017 Spain Populationstatistics from 17AutonomousCommunities (AC)from2006-2014

• ExamineassociationsbetweengenderinequalityandIPVovertheperiodof2006-2014

EcologicalNOTE:correlation

• ReportedIPVcases;• IPVmortality(female)

Gender InequalityIndex (GII) and itscomponents

1. Greatergenderinequalityin2006associatedwithhigherIPVmortalityandhigherIPVreports.

2. In20143. RiskofIPVdeclinedbetween2006and2014.

4. In2014,IPVmortalitywaslowestinareaswithgreatergender

71

inequality.Gressard et al,2015

USA US Staterepresentativedata

• ExamineassociationsbetweenGIIandphysicalandsexualADV

Ecological Physical and sexualADV(female)

Gender InequalityIndex(GII)

GII significantly associatedwith state prevalence offemale physical ADV, butnotfemalesexualADV.

Kim&Kim,2014 148countries Population datafrom 148countriesobtained fromWHO

• EstimateassociatedbetweenPLSE&LFPRcomponentsofGIIandHLE

EcologicalNOTE:correlation

HLE(maleandfemale) PLSE & LFPRcomponents ofGender InequalityIndex(GII)

Gender inequality in theattainment of secondaryeducationandlabourforceparticipation (femaledisadvantage)isassociatedwith reduced healthy lifeexpectancyatbirth.

Mark,2013 96nations Population datafrom Unitednations

• Examineassociationsbetweenfemaleheightandgenderinequality

Ecological Humanheight(female) Gender InequalityIndex(GII)Human developmentindex(HDI)

Years of Education, Life-expectancy, per capitaincome, maternalmortality, maternaldeath, education ratio(females to males,25years and older thathave attained asecondary education),ratio of female to maleseats held in a lower orsinglehouse,oranupperhouse or senate),Childhoodmortalityrate

Gender inequality is mostsignificant factorcontributing to populationheight.

Wellsetal,2012 68countries DatafromObesityTask Forcewebsite

• Examinegendergapinobesityprevalence

• Testwhethereconomicinfluence,andeconomicandgenderinequalityareassociatedwithbetweencountrydifferencesingendergapsinobesityprevalence

Ecological Obesity prevalence(2011)

Gender InequalityIndex(GII)(2005)GINI(2004)PercapitaGDP(2004)

Geographic latitude(estimated as midpointofcountry);totalfertilityrate;age

• Greaterprevalenceofobesityinwomen.

• AssociationbetweenobesityprevalenceandGDP(attenuatingathigherGDP)

• Greaterfemaleexcessinobesitygreaterincountrieswithgreatergenderinequality.

Hassanzedeh et al,2014

123countries Countries in theUnited NationsDevelopmentProgramme (with

• Examinegenderinequalitiesinhealthinrelationtoprevalenceof

Ecological Prevalence of smoking,BloodpressureAlcohol consumption,NCDmortality

Gender InequalityIndex(GII)

• Negativeassociationbetweengenderinequalityand:smoking;alcohol

72

adequatedata) smoking,bloodpressure,alcoholconsumption,cancerdeathrate,BMI

BMICancerLifeexpectancy

consumption;lifeexpectancy,cancer,BMI,bloodpressure

• PositiveassociationbetweengenderinequalityandNCDmortality

Worldeconomicforum:GGIWitvlietetal,2014 32 countries (low

income to highincome but mainlylowincome)

Women aged 18-50in32countries,n=57182participating inthe World HealthSurvey

• Comparethegeneralhealthofsinglemotherswiththatofotherwomen

• Examinewhetherassociationsbetweenhealthandgenderinequalityisstrongeramongsinglemothersthanotherwomen

Cross-sectional surveyandecologicaldata

Self- assessed generalhealth

Global Gender GapIndex

Motherhood status andpartnershipstatusIndividual covariates:age,occupationalstatus,educationalattainmentCountrylevel:GDP

Country gender inequalityis not related to relativehealthoflonemothers

Grittneretal,2012 25 countriesworldwide: mix ofregions

42655 individualsparticipating inGENACIS project,aged25-69

• Examineinfluenceofcountry-levelcharacteristicsandindividualsocio-economicstatusonindividualalcoholrelatedconsequences/harm

Alcohol relatedconsequencesexperienced byindividual(internal)e.g.guilt/remorse; unableto remember nightbefore; failing to dowhat was normallyexpected; unable tostop drinking oncestarted; needing drinkinmorningtogetgoingafterheavydrinkingConsequencesexperienced beyondindividual (external),e.g. effect on finances,effect on marriage,children, getting intofight.

Individual SES: highesteducationallevel

Type of alcoholconsumption (moderate,heavy, monthly, bothheavyandmonthly).Country level covariates:purchasingpowerparity,measure of grossnational income percapita,GINIcoefficient;GenderGapIndex

1. Lowereducatedmenandwomenmorelikelytoreportalcoholrelatedproblems(internalandexternalproblems)thanhighereducatedmen.

2. Lowereducatedwomenmorelikelytoreportexternalalcoholrelatedproblems

There were no significantassociations betweenGender Gap Index andeither internal or externalalcoholrelatedproblems

Van Tuyckomet al,2012

27 Europeancountries

Participants in2005Eurobarometersurvey

• Examinegenderdifferencesinleisuretimephysicalinactivity(LTPI)

Cross-sectional surveyandecologicaldata

Leisure time physicalinactivity(LTPI)

GenderGapIndex AgeMaritalstatusEducationalattainmentUrbanisation

3. GreatergenderequalityassociatedwithreducedgendergapinLTPIdifferences

73

N=25745 • Examinecompositionaleffectsofindividual-levelfactorsoncrossnationaldifferences

• Examinerelationshipsbetweengenderinequalityandgenderdisparities

Employment

–differencedisappearedincountrieswithhighgenderequality.

UNgenderequityindex:GEMVandeVeldeet al,2012

25 countries inEurope

Participants inEuropean Socialsurvey,2006-2007(n=39891) aged18-75years

• Examineassociationsbetweengenderequalityanddepressionamongmenandwomen.

Cross-sectional surveyandecologicaldata

Depression Genderempowermentmeasure(GEM)

GINIGenderAgeAgesquaredEmploymentstatusEducationallevelHouseholdincomeMaritalstatusCohabitationstatusPresence of childrenagedunder12years

1. Highergenderequalityisassociatedwithlowerlevelsofdepressionamongmenandwomen

2. Amongsomesubgroups,somedimensionsofgenderequalityareassociatedwithareducedgenderdifferenceinratesofdepression

Bosque-Prousetal,2015

16 Europeancountries

N=26017 adultsaged 50-64 yearsin SHARE project,2010-2012

• Examinegenderdifferencesinhazardousdrinkingandassessassociationswithcontextualfactorsincludinggenderempowerment

Cross-sectionaldesign Hazardous drinkinglevels

Gender;Standardised genderempowerment (GEM)(examined as acovariate)

Individual: educationallevel, age, migrationstatus, employmentstatus, self perceivedhealthCountry level:standardisedGDP;Standardised genderempowerment(GEM);Standardisedunemploymentrate;Standardised degree ofalcohol advertisingrestrictions;Other standardisedalcoholcontrolpoliciesDrinking patterns scorefor12005;Percentageimmigrants

3. Greatergenderequalityassociatedwithlowergenderdifferencesinhazardousdrinking,withhigherlevelsofalcoholconsumptionamongwomenunderpinningthis.

74

Mayer,2000 26 (GEM) & 37(HDI)countries

1996 WHO datafor26countries

• Examineassociationsbetweensuicideratesandgenderequality

Ecological Age-standardisedsuicide rates (per100000)

Genderempowermentmeasure(GEM)Human developmentindex(HDI)

GDP index, Educationindex, gender-relateddevelopment index andlife-expectancyindex

4. Genderequality,asmeasuredbythegender-relateddevelopmentindexwaspositivelyassociatedwithsuicideratesforbothmenandwomen,butparticularlyforwomen

Shah,2008 55 & 65 countriesworldwide

Data obtainedfrom WHO foryears1991-2003

• Examineassociationsbetweengenderequalityandmaletofemalesexratioofsuiciderates

Male to female ratiofor suicide rates in age65-74 years, and 75years+

Genderempowermentmeasure(GEM)Gender relateddevelopment index(GDI)

5. Genderinequality,asmeasuredbyGEMandGDIwasnotsignificantlycorrelatedwithmaletofemalesexratioofsuicideratesineitheragegroup.

Bondetal,2010 22 countriesworldwide

2006-2007 surveydata fromGENACISprojectAge:18-75years

• Examineassociationsbetweencountrylevelequalityanddrinking,andcomparesexdifferencesacrossdifferentsettings

Cross sectional andecological

Frequency of drinkingin public settings andfrequency of drinkingin private settings inpast12months

Genderempowermentmeasure(GEM)Gender InequalityIndex(GII)GGIeducationGGI politicalparticipation

IndividualAgeGenderMaritalstatusCountrylevelGDP

6. Mendrinkmorethanwomenineachsetting

7. Inadjustedmodelsneithermeasurepredictsgenderdifferencesbutinunadjustedmodelsgreaterreproductiveautonomyandlessviolenceagainstwomenassociatedwithsmallerdifferencesindrinkinginpublicsettings.

GenderDevelopmentindex(GDI)Hopcroft andBradley,2007

29 countries worldwide

1990 survey datafrom WorldValuesSurvey

• Examinesexdifferencesindepression

Ecological and crosssectional

Depression Gender relateddevelopment index(GDI)

IndividualSexAgeSelf-ratedhealthSocioeconomicstatusMaritalstatusPsychologicalstatusCountrylevelPercentMuslim

Rates of depression arehigher in low genderequitycountriesGender gap in depressionlarger in countries withhighgenderequity

75

Notes: US: United States; SES: Socio-economic status; ; LTPA: leisure time physical activity; ; LTPI: leisure time physical inactivity; ISSP: International Social SurveyProgramme; GDP: gross domestic product; SHARE: Survey of Health, Ageing and Retirement in Europe;WHO:World Health Organisation; SRH: Self-rated health; LLI:Limiting long-standing illness;HDI:humandevelopment index;GGS:generationsandgendersurvey;DHS:demographichealthsurvey; IPV: inter-partnerviolence;ADV:adolescentdatingviolence;HLE:healthylifeexpectancy;PLSE:Populationatleastsecondaryeducation;LFPR:labourforceparticipationrate;PSS:perceivedstressscale

76

Appendix

TableA1:Searchterms.Searchiteration

Genderequality/equity

Healthoutcome

Design

1 “genderequality”,“genderequity”,“women”

“morbidity”,“mortality”,“healthbehaviours”,“mentalhealth”

ecological,panelstudy,longitudinaldesign,observational,cohort

2 “genderequality”,“genderequity”,“genderinequality”,“genderinequity”“women”

“health”