Risk-taking behaviour is more frequent in teenage girls with multiple sexual partners

Abstract

Objective. To investigate associations between sexual behaviour and risk-taking health behaviour among adolescent females in our changing sexual culture. Design. A questionnaire study. Girls who had had multiple sexual partners (at least five in their lifetime or four during the last six months) were compared with those with fewer partners. Logistic regression was applied. Setting. The Adolescent Clinic, a primary healthcare unit in the city of Tampere, Finland. Subjects. A sample of 247 female clients aged 15–18 years who had experienced sexual intercourse. Main outcome measures. Contraceptive practices, substance use, and sexual attitudes. Results. Girls with multiple sexual partners (n = 69) and the reference group (n = 178) did not differ from each other significantly by age, age at menarche, or educational status. In univariate analysis, age at sexual debut, contraceptive practices, and various substance uses were strongly associated with having multiple sexual partners. Ever-use of emergency contraception was marginally associated, while ever-use of conventional hormonal contraception or condoms was not. In multivariate analysis, low age at sexual debut (OR 8.75 for age 11–13), omitting contraception at the most recent intercourse (OR 3.48), ever-use of withdrawal as a contraceptive method (OR 2.34), and repeated use of drugs (OR 4.10) were associated with having multiple sexual partners. Conclusion. Different types of risk-taking behaviour are still interlinked. In discussions with adolescents showing one type of risk behaviour health service providers should make an effort to identify other modes of risk-taking.

• Adolescent
• adolescent behaviour
• family practice
• Finland
• Nordic
• reproductive behaviour
• sexual behaviour

The sexual climate has undergone a remarkable change in Western societies during recent decades. From the early 1990s onwards, this phenomenon has been referred to as overeroticization or hypersexualization even in societies like the Nordic with an open-minded sexual culture [1]. Sexual messages have become an established part of everyday culture across the public media [1]. Young people represent a crucial group of consumers of the media and entertainment and pornography [2].

Questions have been raised regarding the effects of such a sexualized culture on adolescent behaviour and sexual health [3]. For instance, studies in the USA have shown that music-video consumers more often engage in risky sexual behaviour [3]. Certain changes in adolescent sexual behaviour have likewise been reported in the Nordic countries, suggesting that adolescents have more sexual partners [4] and that sex without commitment to a relationship has become more common [5]. In Finland, a number of indicators of adolescent sexual behaviour would imply an increase in sexual activity in the late 1990s [6].

In changing sexual culture, it is questionable whether having multiple sexual partners (MP) in adolescence is still a signal of risk-taking behaviour.

• A quarter of consecutive female clients in the adolescent clinic were classified in the MP behaviour group.

• Early age at sexual debut, ineffective contraception, and substance use were associated with MP behaviour.

• Adolescent health service providers should be aware of these interconnections and take them into consideration when counselling adolescents.

So far, having multiple sexual partners in adolescence has been identified as an indicator of the risky lifestyles associated with early sexual debut and unsafe sexual practices [7–11], smoking [12], and substance use [10], [13–15]. Primary healthcare is in a key position to identify adolescents evincing such behaviour and possibly in need of preventive counselling and social support. In the light of the substantial changes in sexual culture and behaviour, an interesting question in clinical work is whether associations between risky sexual behaviour and other modes of risk-taking continue to be valid. Or is it simply the done thing to have multiple sex partners in adolescence nowadays?

The aim of this study was to investigate associations between various aspects of sexual behaviour among adolescent female clients in a primary healthcare setting. We studied the sexual behaviour of girls who reported multiple partners and analysed the associations of this behaviour with sexual and contraceptive history, current contraceptive practices, substance use, and sexual attitudes.

Material and methods

This study was carried out at the Adolescent Clinic of Tampere, the third largest city in Finland with over 200 000 inhabitants. This is the only clinic in Tampere which provides principally sexual health services to adolescents. The study protocol was approved by the Ethics Committees of Tampere City and the Pirkanmaa Hospital District.

Material was collected by the Adolescent Clinic in spring 2004. Nurses introduced the study and gave a self-administered, anonymous questionnaire to 300 consecutive female clients aged 15–18 years. Ten clients refused to participate because they were “too busy”. The participants were asked to complete the questionnaire at the end of a visit and return it to a locked mailbox in the waiting room. The structured questionnaire was based in part on earlier Finnish studies [16] concerning adolescent sexual health and incremented with questions targeting factors associated with risk-taking sexual behaviour [4–6], [8], [9]. Altogether, the questionnaire comprised nine pages and 45 questions, each with 4–10 response alternatives.

Of the 300 questionnaires distributed, 271 were returned completed (response rate 90%). All respondents were urban females with a mean age of 16.75 years (SD 1.08 years). Girls who reported having had at least one coital experience and who responded to the questions on sexual behaviour were included in the study (n = 247).

The participants were divided into two groups according to the reported number of sexual partners. The literature gives no unambiguous criterion regarding a limit for multiple partners, but one figure frequently presented is four or more sexual partners in one's lifetime [17]. Based on previous Finnish and Nordic studies and clinical experience, an even stricter criterion was applied here. Girls who had had four or more sexual partners during the last six months or at least five sexual partners in their lifetime (n = 69) were categorized as the multiple partner (MP) group and the remainder as the reference group (n = 178). Of those who were included in the MP group, 47 (68%) met the criterion of five lifetime partners, four (6%) that of four partners during the last six months, and 18 (26%) both criteria.

Variables for analysis were chosen based on earlier research literature on adolescent risk-taking sexual behaviour [15], [18–20]. First, cross-tabulations were made to study differences between the MP and the reference group (Fisher's exact test or Pearson's chi-squared test). Second, logistic regression was applied to model sexual behaviour in girls with multiple partners. Variables significantly associated with multiple partner behaviour in cross-tabulations were analysed by univariate logistic regression. Finally, variables showing a significant association with multiple-partner behaviour in the univariate model were included in the multivariate logistic regression model. Statistical analyses were performed using SPSS for Windows, version 11.5 (SPSS Inc, Chicago, Illinois, USA). A p-value of less than 0.05 was considered statistically significant.

Results

The MP group and the reference group did not differ from each other significantly by age, age at menarche, or educational career (Table I). Experiences of STI (sexually transmitted infections), childbirth, and miscarriage were more common in the MP group, but the groups did not differ by history of induced abortion.

Table I.  Background data on the study participants.

	Multiple sexual partners group (N = 69)		Reference group (N = 178)		All (N = 247)		p
Age (years)
Mean		17.23		17.37		17.27	0.327
Standard deviation		1.08		1.03		1.07
	%	(n)	%	(n)	%	(n)
School/work attendance
At work	7	(5)	3	(6)	5	(11)
At school or studying	84	(58)	91	(162)	89	(220)
Nothing/Neither/Other	9	(6)	5	(9)	6	(15)
Missing	0	(0)	1	(1)	0	(1)	*0.297
Type of school attended
Vocational	19	(13)	10	(18)	13	(31)
Upper secondary	46	(32)	58	(103)	55	(135)
Comprehensive	16	(11)	19	(33)	18	(44)
Other	3	(2)	3	(6)	3	(8)
Missing	16	(11)	10	(18)	12	(29)	0.208
Menarche age
9–11 years	28	(19)	15	(27)	19	(46)
12–13 years	61	(42)	69	(122)	66	(164)
14–16 years	12	(8)	16	(29)	15	(37)	0.072
Delivery or miscarriage
No	77	(53)	90	(161)	87	(214)
Yes	10	(7)	2	(3)	4	(10)
Missing	13	(9)	8	(14)	9	(23)	0.004
Induced abortion
No	80	(55)	86	(153)	84	(208)
Yes	7	(5)	6	(11)	7	(16)
Missing	13	(9)	8	(14)	9	(23)	0.417
Sexual transmitted disease e
No	86	(59)	98	(175)	95	(234)
Yes	13	(9)	2	(3)	5	(12)
Missing	1	(1)	0	(0)	0	(1)	* < 0.001

*Fisher's exact test; others: Pearson's chi-square.

In univariate analysis a trend was found indicating that the lower the age at first sexual intercourse, the higher was the likelihood of having multiple sexual partners (Table II). Omitting contraception at the most recent intercourse and ever-use of withdrawal as a contraceptive method were both strongly associated with MP sexual behaviour. Ever-use of emergency contraception was marginally associated, whereas ever-use of conventional hormonal contraception or condoms was not (Table II). Furthermore, smoking, binge drinking, and drug use were all significantly associated with sexual behaviour in the MP group (Table II).

Table II.  Odds ratios (with 95% confidence intervals) for adolescent risky sexual behavior, defined as having had multiple sexual partners:* Univariate and multivariate logistic regression analysis.

	Univariate model				Multivariate model
Risk factors	n	OR	95%CI	P	n	OR	95%CI	P
Age	247	1.14	0.87–1.48	0.33	247	1.45	1.00–2.09	0.05
Age sexual initiation (n = 243)
16–18 years	97	1.00			93	1.00
14–15 years	115	3.19	1.58–6.43		108	5.60	2.17–14.46
11–13 years	31	8.95	3.56–22.50	<0.01	30	8.75	2.44–31.38	<0.01
Contraception at the most recent intercourse
Yes	208	1.00			194	1.00
No	38	4.71	2.29–9.69	<0.01	37	3.48	1.32–9.17	0.01
Earlier used contraception methods (n = 247)
Withdrawal
No	169	1.00			157	1.00
Yes	78	3.56	1.94–6.52	<0.01	74	2.34	1.12–4.90	0.02
Emergency contraception
No	145	1.00			133	1.00
Yes	102	1.98	1.08–3.61	0.03	98	2.06	1.00–4.26	0.05
Hormonal contraception
Yes	178	1.00
No	69	0.71	0.37–1.35	0.30
Condom
Yes	236	1.00
No	11	0.24	0.03–1.96	0.18
Consistency of condom use (n = 245)
Never using at all	49	1.00			46	1.00
Using occasionally	163	1.53	0.74–3.16		155	2.10	0.81–5.44
Alwayss using	33	0.31	0.08–1.19	0.03	30	0.85	0.17–4.11	0.16
Smoking (n = 247)
No	115	1.00			104	1.00
Yes	132	3.08	1.68–5.64	<0.01	127	0.94	0.41–2.18	0.88
Binge drinking (n = 237)
Never or a few times a year	133	1.00			128	1.00
Once a month or more often	104	3.17	1.77–5.68	<0.01	103	1.82	0.88–3.78	0.11
Lifetime drug use (n = 247)
No	171	1.00			158	1.00
1–2 times	36	2.99	1.36–6.51		35	1.42	0.54–3.72
3 times or more	40	7.83	3.69–16.61	<0.01	40	4.10	1.58–10.66	0.02
Opinion of probability of getting pregnant if not using contraception (n = 247)
Very probable	133	1.00			125	1.00
Possible	95	1.14	0.63–2.08		87	0.80	0.38–1.73
Very improbable	19	3.37	1.26–9.00	0.05	19	2.68	0.70–10.27	0.23
Opinion of probability of getting STI if not using a condom (n = 247)
Very probable	107	1.00
Possible	128	1.15	0.65–2.04
Very improbable	12	1.41	0.39–5.05	0.82
Attitude towards induced abortion (n = 247)
Alwayss acceptable	73	1.00
Acceptable in certain circumstances	170	1.33	0.70–2.52
Never acceptable	4	9.88	0.96–101.30	0.14

*at least 5 sexual partners in life-time or at least 4 during the last six months.

Among girls reporting multiple partners, sexual behaviour was associated with their views regarding the probability of becoming pregnant without contraception, but not with their opinions regarding the risk of STI without using a condom (see Table II). There was a tendency in the MP group for attitudes towards induced abortion to be less permissive than in the reference group, but the difference was not significant.

Ten variables were entered in the model for multivariate analysis (see Table II), including age as a continuous variable. Of these, age lower than 16 at first sexual intercourse (OR 5.60–8.75), non-use of contraception at the most recent intercourse (OR 3.48), ever-use of withdrawal (OR 2.34), and drug use at least three times in lifetime (OR 4.10) were significantly associated with MP sexual behaviour.

Discussion

This study was carried out among 247 female clients of an adolescent clinic in a primary healthcare setting. We studied factors associated with having multiple sexual partners in adolescence. In multivariate analysis, young age at sexual debut, non-use of contraception at the most recent intercourse, ever-use of withdrawal, repeated drug use, and, marginally, ever-use of emergency contraception were associated with having multiple sexual partners.

We defined multiple sexual partnership as having at least five partners in one's lifetime or having at least four different partners during the last six months. The choice of this cut-off point was based on the variable distribution in the present material and on international research data. The cut-off points used in international literature have varied from three partners in one's lifetime [9] or during the past 12 months [4] to four or more partners during one's lifetime [17]. The appropriate criterion also greatly depends on the cultural context. In view of the wider social acceptance of adolescent sexuality in the Nordic countries compared with most other Western cultures [2], a higher criterion for multiple partnership was chosen here.

MP behaviour was associated with early sexual debut, which is congruent with findings in many earlier studies [7], [9], [10]. In contrast, early menarche was significantly not related to MP behaviour in this study, contrary to many earlier reports of an association between low menarche age and risk-taking behaviour [21]. This finding may suggest that nowadays factors other than physical maturation play a role of growing importance in initiating sexual activity in adolescence.

Neglect of contraception at the most recent intercourse and a history of multiple sexual partners were associated here, as also shown in earlier research [22], [23]. Moreover, MP behaviour was significantly associated with ever-use of the withdrawal method and marginally with ever-use of emergency contraception, which would imply that experiences of unplanned coital events were more frequent in this group. All the results on contraceptive practices clearly support the conception that MP behaviour continues to constitute a signal of risky sexual behaviour.

Regarding participants’ past use of other contraceptives, no difference was seen between the MP group and the reference group with respect to ever-use of hormonal contraception or condoms. In Finland, oral contraception is an established and widely used method of preventing pregnancies in adolescence, which may explain the absence of significant differences between the groups. Access to contraceptive services is easy for adolescents and oral contraceptive use shows little variation in different parts of the country or among different social groups [24].

Binge drinking, smoking, and drug use were associated with MP behaviour in univariate analysis. In multivariate analysis, only drug use entered the model, while smoking and binge drinking dropped out, probably by reason of an interaction with drug use. Smoking and drinking usually precede drug trials and use of drugs alone is less common [25], [26]. Many studies have indicated that various types of risk behaviour are interconnected [18]. Alcohol users have been shown more frequently to fail to use contraception [27], to engage in sex with casual partners [15], or to have multiple sex partners [8]. On the other hand, in some studies the association between risky sexual behaviour and alcohol use [19] or drug use [28] has remained equivocal.

The strong associations found in this study between adolescent sexual behaviour, binge drinking, and drug use have not been previously published in Finland. The findings are extremely interesting in the light of concurrent national trends in substance use and adolescent sexual health. The use of illicit drugs is a fairly new phenomenon in Finnish society, with the steepest increase in the latter half of the 1990s. Binge drinking also increased markedly among adolescents during the same period. Concomitant with these changes, teenage pregnancies and Chlamydia transmissions increased steeply [29], [30]. In the 2000s, alcohol and drug use [31] levelled off and at the same time a levelling off of the indicators for sexual health was noted [32]. The present findings suggest that there is a connection between all these trends and further research on this important issue is clearly warranted.

The response rate (90%) in the present study was high and the questionnaires were carefully completed. The number of participants was, however, rather small and possibly for this reason some differences between the factors studied did not emerge as significant. It is impossible to check whether there is under- or over-reporting, but according to empirical experience and some studies [33], answers can be considered reliable. The data were presumably well representative of clients of this age at the clinic site in question and, with some reservations, of adolescent clients of family planning clinics in other Finnish cities, but not of all adolescents of that age in Finland.

The question in the background of this study was whether having multiple partners has become normative sexual behaviour in adolescence, hence no longer predicting a risk-taking lifestyle in general. Our results do not support such a claim, but clearly suggest that the associations shown in earlier research remain valid under the changing social circumstances. Having multiple sexual partners was associated with risky behaviour such as ineffective contraceptive practices and substance use, and accordingly equally indicates a risk-taking lifestyle in current society. It is important that adolescent health service providers be aware of this interconnectedness so that in discussions with adolescents showing one type of risk behaviour they might make an effort to identify other modes of risk-taking as well.

Acknowledgements

This study was financially supported by the Medical Research Fund of Tampere University Hospital, and the Pirkanmaa Regional Fund of the Finnish Cultural Foundation.

The authors would like to thank Tiina Luukkaala of the University of Tampere for statistical advice, and Robert MacGilleon for language checking.