1. Attachment style and condom use across and within dating relationships
Panos Kordoutis
Panteion University of Social and Political Sciences, Greece
Konstantinos Keskinis
University of Amsterdam, the Netherlands

2. Context and goal
Past and new HIV interventions → not very successful, because
they usually focus on individual characteristics
they are not always based on psychological theories
People are aware of HIV/other STDs and condom’s value, but they don’t consistently use it, especially within close relationships.

3. Context and goal : Two different approaches in research
Individual-specific
It considers condom use decision as an individual decision
It focuses on individual factors like gender, age, age of sexual debut, condom use at sexual debut, attachment styles (an individual- dispositional factor)
Relationship-specific
It considers condom use decision as a common decision between partners
It focuses on relationship factors like relationship type, duration, condom use at first sexual
intercourse in the relationship, partners’ age difference,
coital sex frequency

4. Goal of this study
The different role of individual and relationship factors as predictors of condom use behavior
We investigated this role with four different dependent variables:
Condom use frequency across different relationships
Condom use frequency in a specific relationship
Condom use behavior at the beginning of the relationship (first intercourse)
Most recent condom use behavior in the relationship (last intercourse)

5. Outline
Introduction (attachment theory: extension to adulthood and relation to sexuality – hypotheses)
Method (participants, procedure)
Results (four regression analyses)
Discussion
Conclusions (possible implications in health interventions)

6. Attachment theory Attachment theory (Bowlby, 1969, 1973, 1980)
Children interact with their primary caregivers (usually mothers)
They develop cognitive working models of self and other  affect their later interactions
Attachment theory in adulthood (Hazan & Shaver, 1987)
Adult romantic relationships  many similarities with early attachment relationships
Individual differences in adult attachment’s cognitive working models of self and others  consistent with attachment theory

7. Attachment theory
Individual differences in adulthood  four different attachment styles (Bartholomew & Horowitz, 1991):
Secure (positive self-image and image of others)  desire for relationships, comfort with intimacy, autonomy
Preoccupied (negative self-image and positive image of others)  desire for relationships, fear of rejection
Fearful (negative self-image and image of others)  avoidance of relationships, desire for emotional closeness, fear of rejection
Dismissing (positive self-image and negative image of others)  avoidance of relationships, denial of emotional closeness

8. Attachment styles and condom use
The sexual system is strongly connected with the attachment system
Sexual behavior is connected with partners’ interaction goals in relating
We predict an association between attachment style and condom use behavior:
Condom use requires communication between partners, so high anxiety and fear of rejection may prevent it
Condom non-use may be considered as an indication of trust and higher intimacy

9. Main hypotheses
Individual factors and attachment styles will be strong predictors of condom use behavior during the last 12 months and at the beginning of a specific relationship but not at the most recent intercourse of a relationship or condom use frequency during a relationship
Relationship factors will influence condom use behavior at the level of a specific relationship (most recent intercourse and frequency of condom use).

10. Method 339 participants (men = 158, women = 181)
University and Technical College students
Age mean: 21
They lived in Athens (63%), Thessaloniki (15%), and other cities (22%) in Greece
Participant selection criteria:
Students
Sexually active
Involved in a current or previous but recent (less than 12 months) relationship
duration of relationship  at least one week
53.4% had current, 46,6% referred to a past relationship
Relationship duration
M= days, SD =

11. Materials and procedure
Questions about demographics (gender, age, residence, occupation, family status)
Relationship Questionnaire (RQ, , Bartholomew & Horowitz, 1991)
Four descriptions of attachment patterns
Participants first chose the pattern that was most descriptive of themselves
Participants then rated themselves on a scale ranging from 1 (it absolutely describes me) to 7 (it does not describe me at all)
Questions on sexual behavior (sexual profile)
Past: number of past sexual intercourses, number of past sexual partners, age of sexual debut, condom use at sexual debut
Present: number of sexual partners during the last 12 months

12. Materials and procedure
Questions on the specific relationship:
Present relationships  relationship begginning time
Past relationships  beginning and end
Frequency of sexual intercourse (1= no sexual intercourse, 7= everyday)
Questions about the partner:
Gender, age, education
Open question about the relationship
Free description (a focus manipulation)
Questions about condom use
Frequency during the last 12 months (7-point scale,1= never, 5 = many times, 7 = always)
At the first intercourse of the relationship (Yes/No)
At the most recent intercourse of the relationship (Yes/No)
Frequency during the relationship (7 – point scale)

13. Materials and procedure
Voluntary and anonymous participation
They completed the questionnaire individually
Procedure lasted about 15 minutes
They completed the questionnaire in the university or technical college.

14. Results: 4 regression analyses
Dependent variables:
Condom use frequency during the last 12 months
Condom use at the first intercourse of the relationship
Condom use at the most recent intercourse of the relationship
Condom use frequency during the relationship
Predictors
Individual variables in analysis 1.
Both individual and relationship variables in analyses 2,3,4.

15. Results : 4 regression analyses
Low stress in combination with low avoidance (secure attachment) and condom non-use at sexual debut  less frequent condom use during the last 12 months
The older the age of sexual debut and condom use at sexual debut  greater condom use at the first sexual intercourse of the most recent relationship
Higher coital sex frequency in the relationship, partner’s higher education level, and condom use at the first relationship intercourse  greater condom use at the most recent intercourse of the relationship
Partner’s higher education level and condom use at the first intercourse of the relationship more frequent condom use during the relationship

16. Results : 4 regression analyses
Individual and relationship characteristics
During 12 months
1st intercourse (relationship)
Most recent intercourse (relationship)
During (relationship) β OR
Secure
-.12*
Condom use at sexual debut
.24****
22.54****
Age of sexual debut
1.35*
Coital sex frequency in the relationship
.63***
Partner’s education
1.35***
.11*
Condom use at 1st intercourse of the relationship
7.30****
.35****
****p < *** p < * p < .05

17. Hierarchical regression analysis (criterion variable: condom use during the last 12 months)
Variables Β
SEB β R2
ΔR2
Step 1
.007
Gender
.05
.19
-.02
Age
-.07
-.08
Step 2
.031
.024
Secure
-.11
-.12*
Fearful
.02
Preoccupied
.04
-.09+
Dismissing
.01
Step 3
.098****
.067
Age of sexual debut
.10
.06
Condom use at sexual debut
1.15
.26
.24****
Number of partners during life
Step 4
.000
Number of partners (last 12 months)
-.01

18. Logistic regression analysis (criterion variable: condom use at the first sexual intercourse of the relationship)
Variables Β
SEB
Wald
Gender
.13
.31
.17
Age
.03
.08
.21
Secure
-.13
.09
1.96
Fearful
-.12
2.26
Preoccupied
.05
.07
.37
Dismissing
-.06
.64
Number of partners in life
.04
1.62
Age of sexual debut
.30
5.13*
Condom use at sexual debut
3.12
.41
57.21****
Number of partners (last 12 months)
.02
Partner’s age
-.04
Partner’s education
-.07
.14
.29
Relationship duration
.00
1.99
Coital sex frequency in the relationship
-.20
1.43

19. Logistic regression analysis (criterion variable: condom use at the most recent sexual intercourse of the relationship)
Variables Β
SEB
Wald
Gender
-.29
.24
1.42
Age
-.09
.06
1.79
Secure
-.05
.07
.54
Fearful
.82
Preoccupied
.66
Dismissing
-.04
.59
Number of partners in life
.00
.02
Age of sexual debut
.11
.08
1.70
Condom use at sexual debut
.69
.33
4.37
Number of partners (last 12 months)
-.06
.05
1.32
Partner’s age
-.01
.01
Partner’s education
.30
.10
8.26***
Relationship duration
2.03
Coital sex frequency in the relationship
-.46
.14
10.18***
Condom use at 1st relat. intercourse
1.99
.44
20.55****

20. Hierarchical regression analysis (criterion variable: frequency of condom use during the relationship)
Variables Β
SEB β
Gender
-.20
.22
-.05
Age
-.11
.06
-.10
Secure
-.09
-.08
Fearful
.00
Preoccupied
-.06
.05
-.07
Dismissing
-.03
Number of partners in life
-.00
.02
-.01
Age of sexual debut
.17
.07
.15
Condom use at sexual debut
1.35
.31
.23
Number of partners (last 12 months)
-.04
.04
Partner’s age
.01
Partner’s education
.08
.11*
Relationship duration
Coital sex frequency in the relationship
-.15
.09
Condom use at 1st relat. intercourse
1.95
.33
.35****

21. Discussion
Attachment styles and other individual factors  significant predictors only during the last 12 months and at the first sexual intercourse of the most recent relationship
Relationship factors  significant during the relationship and at the most recent sexual intercourse
It appears that attachment styles and other individual factors predict a general tendency of condom use and relationship factors predict self-protective behavior within the context of a specific relationship.

22. Discussion
Secure style (low stress, low avoidance): although there is strong evidence that it is the most functional style in close relationships and psychological well- being, it may undermine self – protective behavior
condom initiation and use may be perceived as a threat to the intimacy, trust, and increased comfort in the relationship that secure style motivates.
The marginal significance of the preoccupied style effect (high stress, low avoidance) enhances this latter point.

23. Discussion
Condom use at sexual debut  significant predictor of condom use both during the last 12 months and at the first sexual intercourse of the relationship. It may function as a pattern that influences later behavior.
Early age of sexual debut  consistently with previous research, people who engage early in sexual activity appear to be more vulnerable to sexual risk behavior.
Sexual activity in young age may be connected with other factors, such as lower education, less knowledge about HIV/other STDs and less communication with partners.

24. Discussion
Condom use at the first sexual intercourse of the relationship  significant in both frequency and most recent sexual intercourse condom use: it may also function as a pattern forming later sexual behavior and influencing both partners.
Higher educational level of partner  also significant in both cases: based on the high educational level of the sample (university and technical college students), it is likely that the compatible education level of the partner enhances communication and condom use.
Higher coital sex frequency  it makes partners feel more comfortable with intimacy and have an illusion of trust, so it can undermine self-protective behavior.

25. Conclusions 2 different levels of self-protective behavior
General tendency  influenced more by individual and dispositional factors
Within a specific relationship  influenced more by the context of this relationship
Health interventions should
focus on early age
enhance communication and negotiation skills and educate people to distinguish trust and intimacy from self-protective behavior
take into account that many people may consistently use condoms across sexual contacts and relationships, but inconsistently within a specific relationship, because they are influenced by contextual factors