Exploring specific-situational factors of risky sexual behaviors among Chinese college students: a mixed methods study

Abstract

Background

Risky sexual behavior refers to sexual activity that increases the likelihood of STIs, unintended pregnancies, and negative psychosocial outcomes. College students are susceptible to such behaviors. This study aimed to identify proximal situational factors contributing to risky sexual behaviors among Chinese college students and to integrate these with individual-level psychosocial correlates using a sequential explanatory mixed-methods design.

Methods

In phase 1 (cross-sectional, September–December 2022), college students from three universities in Daqing, Northeast China, completed an online questionnaire assessing attitudes toward sexuality, perceived social support, protection self-efficacy, sexual assertiveness, and sexual limit-setting self-efficacy. Univariate and binary logistic regression analyses identified factors associated with risky sexual behaviors. In phase 2 (March–July 2023), purposeful sampling selected students who had engaged in risky sexual behaviors for semi-structured interviews. Data were analyzed using NVivo 11.0.

Results

Among 1,761 participants, 25.72% reported sexual activity and 9.60% reported risky sexual behavior; among sexually active students, the prevalence was 37.3%. Specific-situational factors included material environment (economic status, resource accessibility), social environment (partner characteristics, peer influence, entertainment venues), antecedent states (attitudes toward premarital sex, sexual desire/impulse, negative emotions), objective elements (psychological needs, sexual mental health, social media for dating), and opportunity elements (relationship status, solitude).

Conclusions

This study integrates individual-level psychosocial correlates with proximal situational factors to understand risky sexual behaviors among college students. The findings extend individual-level explanations by identifying material, social, emotional, motivational, and opportunity-related contexts that immediately precede risky behaviors, providing evidence for situational risk screening and targeted prevention.

Background

Risky sexual behavior, also referred to as high-risk sexual behavior, is commonly defined as any sexual behavior that increases an individual’s vulnerability to adverse sexual and reproductive health outcomes, including sexually transmitted infections (STIs), HIV infection, unintended pregnancy, and related psychosocial consequences [1]. Although definitions vary across studies, risky sexual behaviors generally include inconsistent or non-use of condoms, multiple sexual partners, casual sexual encounters, and sexual activity under the influence of alcohol or psychoactive substances [1, 2]. These behaviors have become a major global public health concern. According to the World Health Organization, more than 1 million curable STIs are acquired every day worldwide, and an estimated 374 million new infections with chlamydia, gonorrhoea, syphilis, or trichomoniasis occurred in 2020 [3]. Young people are also vulnerable to HIV infection; UNICEF estimated that 370,000 young people aged 15–24 years were newly infected with HIV in 2024 [4]. These global data highlight the importance of preventing risky sexual behaviors during late adolescence and emerging adulthood, a developmental period that overlaps with university life.

College and university students are particularly susceptible to risky sexual behaviors because they experience increased autonomy, separation from parental supervision, peer influence, romantic and sexual exploration, alcohol use, and exposure to digital or entertainment-based social environments. International evidence from regional studies and meta-analyses indicates that risky sexual behaviors are common among university students, although prevalence estimates vary substantially according to cultural context, sample characteristics, and operational definitions. A systematic review and meta-analysis of college and university students in Ethiopia reported a pooled prevalence of risky sexual behaviors of 41.62% [5]. A recent study among Injibara University students in Northwest Ethiopia reported that 38.2% of students engaged in risky sexual behavior and identified peer pressure, substance use, pornography exposure, HIV knowledge, and HIV risk perception as associated factors [6].

Evidence from other regions similarly demonstrates substantial variation in risky sexual behaviors among university students. In Sri Lanka, a cross-sectional study of 1,575 undergraduates from state universities in the Western Province reported that 12.4% had engaged in risky sexual behaviors in the past year, with alcohol consumption and nightclub attendance independently associated with increased risk [7]. In Brazil, a study among freshmen undergraduate students found that the prevalence of risky sexual behavior was 9.0%, and that risky sexual behavior was associated with male gender, alcohol consumption, psychoactive substance use before last sexual intercourse, and use of smartphone applications for sexual purposes [8]. In Portugal, a study of 1,946 university students found that only 39.4% reported consistent condom use, indicating that inconsistent condom use remains common even in higher-education settings [9].

In China, a large-scale survey based on the National College Student Survey on Sexual and Reproductive Health in 2020 showed that 60.59% of college students with previous heterosexual experience had engaged in at least one risky sexual behavior; among them, 41.87% did not use contraception every time they had sexual intercourse, 40.96% had used ineffective or unsafe contraceptive methods, and 13.16% had engaged in casual sex or sex with multiple partners [10]. Another study among 4,769 Chinese female undergraduates found that 18.1% had experienced sexual intercourse; among those sexually experienced students, 17.5% had experienced pregnancy and 22.7% had contracted reproductive tract infections [11]. Taken together, international and Chinese evidence indicates that risky sexual behavior among university students is a heterogeneous and context-dependent phenomenon. Therefore, beyond examining stable individual-level characteristics, it is necessary to explore the specific situational factors that immediately precede risky sexual behaviors among Chinese college students.

These behaviors are associated with adverse physical, psychological, and social consequences among university students. Although a substantial body of literature has examined the role of individual and social determinants in shaping sexual risk behaviors, most studies have primarily focused on relatively stable individual-level factors, such as demographic characteristics, sexual knowledge, and attitudes. Theoretical frameworks such as Social Cognitive Theory [12] and the Information–Motivation–Behavioral Skills Model [13] provide useful lenses for understanding the determinants of sexual risk behaviors. These models conceptualize sexual risk-taking as the result of dynamic interactions among cognitive, social, motivational, and behavioral capability factors. Cognitive and motivational factors, such as attitudes toward sexuality and attitudes toward premarital sex, influence individuals’ beliefs and evaluations regarding sexual behavior. Social factors, including perceived social support, peer norms, and partner influence, can either reinforce or mitigate risk by shaping contextual expectations and access to protective resources. Behavioral capability factors, particularly sexual self-efficacy, including Sexual Limit-setting Self-efficacy includes 3 items, sexual assertiveness, and limit-setting ability, determine whether individuals are able to translate knowledge and motivation into protective behaviors.

Importantly, these psychosocial determinants do not operate in isolation; rather, their effects may be activated, weakened, or reshaped by proximal situational factors. Situational factors refer to the immediate material, social, emotional, and opportunity-related contexts in which sexual decisions occur. Evidence has shown that substance use is associated with risky sexual behaviors among adolescents and young people [14], and alcohol use has been linked to risky sexual behaviors among college students and youth [15]. Event-level studies further suggest that alcohol use before or during sexual encounters may reduce protective behaviors, particularly condom use, in specific sexual situations [16, 17]. However, existing research has often emphasized substance-related or individual-level predictors, while less attention has been paid to the broader situational conditions that immediately precede risky sexual behaviors among college students, such as condom accessibility, partner and peer pressure, emotional distress, online partner-seeking, unfamiliar environments, and relationship dynamics.

Therefore, further research is needed to systematically examine how situational factors contribute to risky sexual behaviors among college students, particularly within the real-world contexts in which sexual decisions occur. To address this gap, the present study used a sequential explanatory mixed-methods design to investigate both individual-level psychosocial correlates and proximal situational triggers of risky sexual behaviors among Chinese college students. This study contributes to the literature in three ways. First, it extends existing individual-level explanations by linking relatively stable psychosocial factors, including sexual attitudes, perceived social support, and sexual self-efficacy, with immediate situational contexts preceding risky sexual behaviors. Second, it applies a situational perspective to organize material, social, emotional, motivational, and opportunity-related factors in sexual decision-making. Third, by integrating quantitative predictors with qualitative narratives, this study provides a more context-sensitive understanding of how risky sexual behaviors occur among Chinese college students and offers a basis for developing situational risk assessment and targeted prevention strategies.

Methods

Study design and setting

This study utilized a sequential explanatory mixed-methods design and was conducted in Daqing City, northeastern China [18]. This design was chosen because the quantitative phase could identify factors statistically associated with risky sexual behaviors, whereas the qualitative phase could further explain how these factors operated in specific situations preceding risky sexual behaviors. In the first phase, 1,761 undergraduates completed valid questionnaires from September to December 2022, with the aim of assessing risky sexual behaviors and associated factors among college students.

In the second phase, 19 students who had reported engaging in risky sexual behaviors in the past year during phase 1 were purposefully selected using maximum variation sampling to ensure diversity in age, gender, grade, major, type of risky sexual behavior, and sexual orientation. These participants were recruited based on the inclusion criteria for the qualitative interviews. They participated in semi-structured interviews from March to July 2023 to explore in depth the situational factors influencing risky sexual behaviors among college students. Data saturation was assessed iteratively during the interview process. After each interview, two independent coders reviewed the transcript and discussed whether new themes or subthemes had emerged. Recruitment continued until two consecutive interviews yielded no new information, and both coders confirmed that thematic saturation had been reached.

Operational definitions of key variables

For the purposes of this study, sexual activity was operationally defined as any form of sexual intercourse, including vaginal, anal, or oral intercourse, that had occurred during the participant’s college years.

Risky sexual behavior (RSB) was operationally defined as engaging in any one or more of the following behaviors during the past year: (1) inconsistent or non-use of condoms during sexual intercourse; (2) having sexual relations with multiple sexual partners, defined as two or more partners; (3) engaging in casual sexual encounters, defined as sexual activity with a non-romantic partner without an established relationship; (4) sexual activity under the influence of alcohol, stimulant drugs, or other illicit/psychoactive substances; (5) engaging in commercial sex, defined as transactional sex; or (6) group sexual activity [1, 2, 10]. Participants who reported at least one of the above six behaviors occurring within the past year were classified as having engaged in risky sexual behavior.

Classification as a case required endorsement of at least one of the six behaviors within the past year. This past-year timeframe was selected to minimize recall bias and ensure the recency of behaviors. Students who did not report any of these behaviors were classified as not having engaged in risky sexual behavior.

Participants

Quantitative research

A multistage cluster sampling procedure based on conveniently selected universities was used in the quantitative phase. Three universities in Daqing City, Heilongjiang Province, northeastern China, were selected according to institutional accessibility and type. Within each university, classes were stratified by grade and randomly selected as clusters. All eligible students in the selected classes were invited to participate in the survey.

The inclusion criteria were as follows: (1) being enrolled as a full-time college student and (2) being willing to participate in the survey. The exclusion criteria included students who had been screened for mental health issues and diagnosed by a psychiatrist with mental disorders such as schizophrenia, major depressive disorder, or bipolar disorder. These students were excluded because severe mental illness may impair judgment and decision-making capacity, potentially confounding the relationship between psychosocial factors and sexual behavior. The sample size was estimated using the formula for estimating a population proportion: N = (Zα/δ)² × p(1 − p), with δ = 2% and α = 0.05. Previous studies suggested that the prevalence of sexual behavior among college students ranged from 15% to 33.67% [19, 20]. After accounting for a 10% non-response rate, the final required sample size was estimated to be between 1,346 and 2,359.

Qualitative research

Students who had engaged in risky sexual behavior were selected from phase 1 using purposeful sampling and maximum variation sampling. Variation in type of risky sexual behavior, age, grade, gender, major, and sexual orientation was considered to enhance the diversity of the qualitative sample and to capture a wide range of situational experiences.

The inclusion criteria were as follows: (1) being a full-time college student; (2) having engaged in risky sexual behaviors during the past year; (3) being able to clearly recall and articulate relevant experiences; and (4) being willing to participate in the interview. The sample size was determined by data saturation, which was indicated by the emergence of no new themes or subthemes during the iterative analysis.

Measurements

Demographic information

A comprehensive questionnaire was used to collect demographic information, including gender, ethnicity, age, grade, major, sexual orientation, family type, parents’ education level, and parents’ occupation. It also captured information on risky sexual behavior based on the operational definition, including the specific types of risky behaviors engaged in, such as condom use, number of sexual partners, and alcohol, stimulant drug, or other psychoactive substance use before or during sexual activity. In addition, the questionnaire collected information on sources of sexual partners, such as classmates, friends, and online platforms, and places where these behaviors occurred, such as hotels, dormitories, and entertainment venues.

Attitude toward sexuality

This scale was included to assess the cognitive/motivational component of the Social Cognitive Theory and Information–Motivation–Behavioral Skills frameworks, reflecting individuals’ evaluative judgments toward sexual behavior [12, 13]. The scale was developed by Yang et al. as the attitudinal subscale of the Sexual Mental Health Questionnaire for College Students and consists of 10 items designed to evaluate attitudes toward sexuality [21]. It assesses individuals’ beliefs, feelings, and evaluative judgments related to sexual behaviors, sexual norms, and personal sexual experiences. Items include statements such as “The decision to have an abortion rests solely with the pregnant woman” and other items assessing attitudes toward premarital sex, sexual morality, and reproductive autonomy. Responses are recorded on a 5-point Likert scale, ranging from 1 = strongly disagree to 5 = strongly agree, with total scores ranging from 10 to 50. Higher scores indicate attitudes that are more conservative, traditional, or conflicted regarding sexuality, which the scale developers described as reflecting more negative sexual mental health [21]. Such attitudes and related negative affect may be associated with increased sexual risk, including unprotected sex [22]. In this study, Cronbach’s α for this scale was 0.72.

Social support rating scale for college students

This scale was included to assess the social-environmental component, capturing both protective and risk-promoting interpersonal influences. The Social Support Scale for University Students was developed by Ye and Dai to assess social support among university students [23]. It consists of 17 items across three subscales: subjective support, objective support, and support utilization. Each item is rated on a 5-point Likert scale, ranging from 1 = very inconsistent to 5 = very consistent. Total scores range from 17 to 85, with higher scores indicating higher levels of perceived social support. In this study, Cronbach’s α for the total social support scale was 0.983, indicating very high internal consistency. The Cronbach’s α coefficients for the three subscales—subjective support, objective support, and support utilization—were 0.925, 0.938, and 0.923, respectively. Given the extremely high alpha for the total scale, the possibility of item overlap or redundancy should be considered when interpreting the results. Nevertheless, because the scale has been validated among university students, the total score was used in subsequent analyses.

Sexual health indicators scale

These indicators were included to assess the behavioral capability component, operationalizing individuals’ perceived self-efficacy in executing protective sexual behaviors [12, 13]. The Sexual Health Indicators Survey was originally developed and validated by Smylie et al. among young people aged 16–24 years [24]. In the present study, three subscales were adopted to assess sexual self-efficacy: Protection Self-efficacy, Sexual Assertiveness, and Sexual Limit-setting Self-efficacy. Protection Self-efficacy includes 9 items, such as “I feel confident in my ability to use protection for myself and/or my partner.” Sexual Assertiveness includes 5 items, such as “I am assertive about the sexual aspects of my life.” Self-efficacy includes 3 items, such as “I feel confident in my ability to choose when and where to engage in sexual activity.” Each item is rated on a 5-point Likert scale, ranging from 1 = strongly disagree to 5 = strongly agree, with higher scores indicating greater sexual self-efficacy and a more positive and responsible approach to sexual health. In this study, Cronbach’s α values for these three subscales were 0.926, 0.706, and 0.722, respectively.

Data collection

Quantitative research

In the quantitative phase, classes within each grade of the participating universities were numbered, and a random-number method was used to select classes as clusters. All students in the selected classes who met the eligibility criteria were invited to complete the anonymous online questionnaire through Questionnaire Star. Surveys were conducted during recess, evening self-study sessions, or class meetings in classrooms. If participants had questions, the first author provided standardized explanations. Electronic informed consent was obtained from all participants before the survey. A total of 2,149 questionnaires were distributed, and 1,761 valid questionnaires were retained, yielding an effective response rate of 81.9%.

Qualitative research

One-on-one, face-to-face semi-structured interviews were conducted to explore the situational factors influencing risky sexual behavior among college students. The interview guideline was developed based on Belk’s situational theory, which emphasizes the role of immediate situational contexts in shaping behavior [25]. The interview guide included the following questions: (1) How familiar are you with the concept of risky sexual behavior, and can you share what it means to you personally? (2) What are your thoughts and feelings about engaging in risky sexual behaviors? (3) Could you describe the detailed process when you engaged in risky sexual behavior, including the environment, purpose, self-status, and opportunity? Could you describe more details to help us understand the context and your experience fully? (4) Are there any other aspects related to risky sexual behavior that you think are important and would like to discuss further?

Interviews were conducted in private, quiet, and comfortable settings, such as a private study room or other enclosed space where conversations could not be overheard. At the beginning of each interview, the researcher explained the purpose of the interview and emphasized the strict confidentiality of personal information. Written informed consent was obtained from all participants. Each interview was audio-recorded and lasted approximately 45–60 min.

Because the interview topics involved sensitive sexual experiences, additional safeguards were adopted. Participants could skip any question, pause or terminate the interview, and withdraw at any time without consequences. Interviews were conducted by researchers trained in qualitative interviewing and sexual health research. If emotional distress or disclosures of coercion or unwanted sexual experiences occurred, the interviewer provided supportive responses and referral information for university counseling services in accordance with the approved ethics protocol. All audio recordings and transcripts were anonymized, stored securely, and accessed only by the research team.

Statistical analysis

Quantitative research

All quantitative data analyses were conducted using IBM SPSS version 26 for Windows. Descriptive statistics were used to summarize participant characteristics and the prevalence of risky sexual behaviors. Chi-square tests and Mann–Whitney U tests were used for comparative analyses. Factors associated with risky sexual behavior were analyzed among sexually active students. Variables significant in univariate analyses were entered into the binary logistic regression model using a stepwise selection procedure. A significance level of α = 0.05 was adopted for all statistical tests.

Qualitative research

Data organization and analysis were conducted using NVivo 11.0 (QSR International, Melbourne, Australia). Based on qualitative content analysis procedures [26], two independent researchers coded the same dataset. The coding process followed four steps. First, the researchers reviewed the transcripts multiple times to gain a comprehensive understanding of the content. Each interview was segmented by question to facilitate detailed analysis, and NVivo 11.0 was used to organize the data into analytical units. Second, during the open coding phase, each interview was analyzed line by line and sentence by sentence to identify meaningful words, phrases, and concepts relevant to the research questions. Third, after open coding, similar codes were grouped into broader themes and then further divided into subthemes to provide a more nuanced understanding of the data. Fourth, to ensure consistency and reliability, two independent researchers compared their coding results. In cases of disagreement, a third expert with experience in qualitative research and sexual health was consulted to resolve discrepancies.

After separate quantitative and qualitative analyses, findings were integrated using a joint display approach, which is recommended for bringing quantitative and qualitative findings together and generating mixed-method inferences [27]. Statistically significant quantitative predictors were compared with qualitative themes and subthemes to identify convergence, expansion, and divergence between the two phases.

Results

Quantitative results

Among the 1,761 participants, 453 college students (25.72%) reported engaging in sexual activity as defined above. The overall prevalence of risky sexual behavior across all participants was 9.60% (169/1,761). However, among those with sexual experience (n = 453), the prevalence was markedly higher at 37.3% (169/453). The most frequently reported risky behavior was sexual activity under the influence of alcohol or other substances. Specifically, 24.06% of sexually active students reported sexual activity after consuming alcohol, and 7.70% reported stimulant use before or during sexual activity. Other risky behaviors included inconsistent or non-use of condoms (23.40%), multiple sexual partners (16.11%), casual sexual encounters (14.13%), group sexual activity (9.10%), and commercial sexual activity (7.50%). Detailed in Table 1.

Table 1.

Sociodemographic characteristics of the study participants (N = 1761)

Variables	Categories	Phase1 (N = 1761, %)	Phase2 (N = 19, %)
Age(M ± SD)		19.20 ± 1.46	20.26 ± 1.86
Gender	Male	506(28.73%)	13(68.42%)
	Female	1255(71.27%)	6(31.58%)
Grade	Freshman	847(48.10%)	1(5.26%)
	Sophomore	399(22.66%)	4(21.05%)
	Junior	309(17.55%)	8(42.11%)
	Senior	206(11.69%)	6(31.58%)
Specialty	Medicine	1173(66.61%)	4(21.05%)
	Non-Medicine	588(33.39%)	15(78.95%)
Monthly living expenses (Yuan)	≤ 1000	165(9.37%)	1(5.26%)
	1001–2000	1235(70.13%)	10(52.64%)
	2001–3000	296(16.81%)	4(21.05%)
	> 3000	65(3.69%)	4(21.05%)
Being an only child	No	835(47.42%)	9(47.37%)
	Yes	926(52.58%)	10(52.63%)
Living place	City	952(54.06%)	8(42.11%)
	Rural	809(45.94%)	11(57.89%)
Type of family	Nuclear family	1082(61.44%)	5(26.32%)
	Joint family	419(23.79%)	6(31.58%)
	Reconstituted family	70(3.98%)	3(15.79%)
	Single parent family	190(10.79%)	5(26.31%)
Monthly income of family (Yuan)	< 3000	390(22.15%)	1(5.26%)
	3000–7500	806(45.77%)	10(52.64%)
	7501–12,000	363(20.61%)	4(21.05%)
	> 12,000	202(11.27%)	4(21.05%)
Father’s education	Low level(Primary and Junior school)	1096(62.24%)	5(26.32%)
	Middle level(Senior school & Junior college)	524(29.76%)	10(52.63%)
	High level (University and above)	141(8.00%)	4(21.05%)
Mother’s education	Low level(Primary and Junior school)	1163(66.04%)	6(31.58%)
	Middle level(Senior school & Junior college)	469(26.63%)	10(52.63%)
	High level (University and above	129(7.33%)	3(15.79%)
Father’s occupation	Civil servant	77 (4.37%)	2(10.53%)
	Freelancer	492 (27.94%)	5(26.32%)
	Professionals	110 (6.24%)	2(10.52%)
	Service worker	70 (3.98%)	5(26.32%)
	Farmer	794 (45.09%)	3(15.79%)
	Others	218(12.38%)	2(10.52%)
Mother’s occupation	Civil servant	49 (2.78%)	3(15.79%)
	Freelancer	566 (32.14%)	3(15.79%)
	Professionals	115(6.53%)	2(10.53%)
	Service worker	144(8.17%)	6(31.58%)
	Farmer	603 (34.24%)	3(15.79%)
	Others	284 (16.14%)	2(10.52%)
Monthly entertainment venue visit	None	768(43.61%)	1(5.26%)
	One time	408(23.17%)	4(21.05%)
	Two times	207(11.76%)	5(26.32%)
	Three times	114(6.47%)	5(26.32%)
	More than four times	264(14.99%)	4(21.05%)
Daily social media usage	Less than one hour	556(31.57%)	8(42.11%)
	1 to 4 h	810(46.00%)	8(42.11%)
	5 to 8 h	235(13.34%)	2(10.53%)
	More than 8 h	160(9.09%)	1(5.25%)
Relationship status	In love	506(28.73%)	11(57.89%)
	Single	1255(71.27%)	8(42.11%)
Attitude towards premarital sex	Agree	305(17.32%)	14(73.68%)
	Neutrality	1059(60.14%)	3(15.79%)
	Disagree	397(22.54%)	2(10.53%)
Discussion of sexual topics with other	Never	563(31.97%)	1(5.26%)
	Few	848(48.16%)	7(36.84%)
	Sometimes	286(16.24%)	7(36.84%)
	Often	64(3.63%)	4(21.06%)
Sexual orientation	Homosexual	69(3.92%)	5(26.32%)
	Heterosexuality	1452(82.45%)	12(63.16%)
	Bisexual	102(5.79%)	2(10.52%)
	Uncertain	138(7.84%)	0(0%)

Univariate analysis

Factors associated with risky sexual behavior were analyzed among sexually active students. The Chi-square test revealed significant associations between risky sexual behaviors and a range of sociodemographic, behavioral, and attitudinal factors, including gender, monthly living expenses, mother’s education and occupation, frequency of entertainment venue visits, daily social media usage, attitudes toward premarital sex, discussions of sexual topics with others, and sexual orientation (p < 0.05). These results, detailed in Table 2, highlight the diverse range of variables influencing risky sexual behaviors.

Table 2.

Chi-square test results on the association between the social-demographic predictor categories and risky sexual behaviors(N = 453)

Variables	Categories	Had risky sexual behaviors		χ 2	p
		No (n = 284)	Yes (n = 169)
Gender	Male	117(41.20%)	106(62.72%)	19.640	0.000**
	Female	167(58.80%)	63(37.28%)
Grade	Freshman	82(28.87%)	58(34.32%)	1.990	0.574
	Sophomore	62(21.84%)	34(20.12%)
	Junior	82(28.87%)	49(28.99%)
	Senior	58(20.42%)	28(16.57%)
Specialty	Medicine	175(61.62%)	95(56.21%)	1.286	0.257
	Non-medicine	109(38.38%)	74(43.79%)
Monthly living expenses (Yuan)	≤ 1000	20(7.04%)	12(7.10%)	14.534	0.002**
	1001–2000	173(60.92%)	92(54.44%)
	2001–3000	79(27.82%)	41(24.26%)
	> 3000	12(4.22%)	24(14.20%)
Being an only child	Yes	145(51.06%)	101(59.76%)	3.237	0.072
	No	139(48.94%)	68(40.24%)
Living place	City	159(55.99%)	102(60.36%)	0.828	0.363
	Rural	125(44.01%)	67(39.64%)
Type of family	Nuclear family	161(56.69%)	109(64.50%)	2.731	0.435
	Joint family	70(24.65%)	34(20.12%)
	Reconstituted family	17(5.98%)	9(5.32%)
	Single parent family	36(12.68%)	17(10.06%)
Monthly income of family (Yuan)	< 3000	65(22.89%)	35(20.71%)	6.102	0.107
	3000–7500	110(38.73%)	51(30.17%)
	7501–12,000	66(23.24%)	45(26.63%)
	> 12,000	43(15.14%)	38(22.49%)
Father’s education	Low level (primary school and below)	176(61.97%)	93(55.03%)	5.215	0.074
	Middle level (Senior school & junior college)	86(30.28%)	52(30.77%)
	High level (University and above)	22(7.75%)	24(14.20%)
Mother’s education	Low level (primary school and below)	192(67.61%)	94(55.62%)	10.291	0.006*
	Middle level (Senior school & junior college)	70(24.65%)	47(27.81%)
	High level (University and above)	22(7.74%)	28(16.57%)
Father’s occupation	Civil servant	18(6.34%)	12(7.10%)	9.027	0.108
	Freelancer	75(26.41%)	51(30.18%)
	Professionals	15(5.28%)	14(8.28%)
	Service worker	15(5.28%)	11(6.51%)
	Farmer	134(47.18%)	57(33.73%)
	Others	27(9.51%)	24(14.20%)
Mother’s occupation	Civil servant	13(4.58%)	8(4.73%)	17.722	0.003**
	Freelancer	87(30.63%)	55(32.55%)
	Professionals	16(5.63%)	21(12.43%)
	Service worker	21(7.40%)	16(9.47%)
	Farmer	109(38.38%)	37(21.89%)
	Others	38(13.38%)	32(18.93%)
Monthly entertainment venue visit	None	93(32.75%)	37(21.89%)	37.395	0.000**
	One time	60(21.13%)	19(11.24%)
	Two times	42(14.79%)	12(7.10%)
	Three times	23(8.10%)	19(11.24%)
	More than four times	66(23.23%)	82(48.53%)
Daily social media usage	Less than one hour	60(21.13%)	32(18.94%)	11.931	0.018*
	1–4 h	130(45.77%)	55(32.54%)
	5 to 8 h	49(17.25%)	38(22.49%)
	More than 8 h	45(15.85%)	44(26.03%)
Relationship status	In love	197(69.37%)	95(56.21%)	8.001	0.005**
	Single	87(30.63%)	74(43.79%)
Attitude towards premarital sex	Agree	109(38.38%)	101(59.76%)	20.985	0.000**
	Neutrality	169(59.51%)	63(37.28%)
	Disagree	6(2.11%)	5(2.96%)
Discussion of sexual topics with other	Never	53(18.66%)	17(10.06%)	44.769	0.000**
	Few	147(51.76%)	63(37.28%)
	Sometimes	71(25.00%)	48(28.40%)
	Often	13(4.58%)	41(24.26%)
Sexual orientation	Homosexual	15(5.28%)	15(8.88%)	12.097	0.007**
	Heterosexuality	237(83.45%)	118(69.82%)
	Bisexual	19(6.69%)	24(14.20%)
	Uncertain	13(4.58%)	12(7.10%)

* p < 0.05, ** p < 0.01

Mann-Whitney tests further identified significant differences in psychosocial variables, including attitudes toward sexuality, perceived social support, and dimensions of sexual self-efficacy such as protection self-efficacy and sexual limit-setting self-efficacy (p < 0.05). These findings suggest that psychosocial factors play a critical role in shaping sexual behavior outcomes (Table 3).

Table 3.

Mann–Whitney test results comparing sexual and psychosocial factors by risk behavior group (N = 453)

Variables	Had risky sexual behaviors, Median(P25, P75)		Z	p
	No (n = 284)	Yes(n = 169)
Attitude toward sexuality	19.0 (14.0,24.0)	22.0 (16.0,30.0)	-4.191	0.000**
Perceived social support	68.0 (55.3,81.0)	67.0 (51.0,74.0)	-2.385	0.017*
Protection self-efficacy	36.0 (29.0,41.0)	35.0 (27.0,39.5)	-2.224	0.026*
Sexual assertiveness	18.0 (15.0,21.0)	19.0 (15.0,23.0)	-0.578	0.564
Sexual limit-setting self-efficacy	12.0 (9.0,15.0)	12.0 (9.0,14.0)	-2.019	0.043*

* p<0.05, ** p<0.01

Multivariate analysis

A binary logistic regression was used to examine factors associated with risky sexual behavior among college students. Univariate analysis through Chi-square tests identified several significant sociodemographic variables, including gender, living expenses, mother’s education and occupation, social media usage, relationship status, and attitudes toward premarital sex (Table 2). Mann-Whitney tests (Table 3) revealed significant differences in attitudes toward sexuality, perceived social support, and self-efficacy related to sexual behavior. These significant variables were then included in the binary logistic regression model to further investigate the factors influencing risky sexual behavior.

The binary logistic regression analysis provided deeper insights into the predictors of risky sexual behaviors (Table 4). Only variables and categories retained in the final stepwise model are shown. Gender emerged as a significant factor, with male students exhibiting significantly higher odds of engaging in such behaviors compared to females (OR = 2.414, p < 0.01). Family background also influenced outcomes, with students whose mothers were farmers were less likely to engage in risky sexual behaviors than those whose mothers were civil servants (OR = 0.490, p < 0.05). Also, behavioral and attitudinal factors significantly contributed to the risk, with students who visited entertainment venues more than four times per month (OR = 2.184, p < 0.01), spent 4–8 h daily on social media (OR = 1.991, p < 0.05), agreed with premarital sex (OR = 1.829, p < 0.05), or frequently discussed sexual topics with others (OR = 4.011, p < 0.01) exhibited greater odds of engaging in risky sexual behaviors. Furthermore, psychological factors also played a notable role, students with more negative sexual mental health, as indicated by higher scores on attitudes toward sexuality, were slightly more likely to engage in risky behaviors (OR = 1.035, p < 0.05). This may be explained by the fact that students with higher scores tend to hold more conflicting, negative, or distorted views toward sexuality, which refers to individuals’ beliefs, feelings, and evaluations related to sexual behavior, norms, and personal experiences. Such attitudes are often associated with greater psychological distress, increased behavioral risk, and challenges in sexual adjustment [21, 22]. In contrast, higher protection self-efficacy was associated with a significant reduction in risky sexual behaviors (OR = 0.951, p < 0.01). Protection self-efficacy refers to an individual’s confidence in their ability to carry out protective sexual behaviors, such as consistently using condoms, avoiding high-risk scenarios, and clearly communicating personal boundaries with a partner [12, 24]. This confidence plays a critical psychological role in influencing behavior. Students with greater protection self-efficacy are more likely to take proactive steps to safeguard their sexual health, underscoring the importance of self-efficacy as a protective factor in sexual decision-making.

Table 4.

Binary logistic regression analysis on risky sexual behaviors in college students (N = 453)

Variables	B	S.E.	Wald	p	OR (95% CI)
Gender
Female (Reference)
Male	0.881	0.236	13.971	0.000**	2.414(1.521–3.832)
Mother’s occupation
Civil servant (Reference)
Farmer	-0.714	0.255	7.843	0.005**	0.490(0.297–0.807)
Monthly entertainment venue visit
None (Reference)
More than four times	0.781	0.240	10.555	0.001**	2.184(1.363–3.498)
Daily social media usage
1–4 h (Reference)
5–8 h	0.689	0.276	6.242	0.012*	1.991(1.160–3.418)
Attitude towards premarital sex
Disagree (Reference)
Agree	0.604	0.234	6.674	0.010*	1.829(1.157–2.891)
Discussion of sexual topics with other
Never (Reference)
Often	1.389	0.382	13.195	0.000**	4.011(1.896–8.488)
Attitude toward sexuality	0.035	0.015	5.284	0.022*	1.035(1.005–1.066)
Protection self-efficacy	-0.050	0.014	12.424	0.000**	0.951(0.925–0.978)
Constant	-0.279	0.660	0.178	0.673	0.757

OR odds ratio, CI confidence interval

 * p < 0.05, ** p < 0.01

These findings, summarized in Tables 2, 3 and 4, highlight the multifaceted nature of risky sexual behaviors among college students. They underscore the complex interplay of sociodemographic, behavioral, attitudinal, and psychological factors.

Qualitative results

For the qualitative phase, 19 undergraduate students who had experienced risky sexual behaviors were invited for in-depth interviews. Content analysis revealed 5 themes and 14 subthemes related to risky sexual behaviors among college students.

Theme 1: material acquisition or use

Subtheme: not easy to access the condoms

Some students indicated difficulty accessing condoms, particularly at residential hostels or friends’ homes. For instance, one participant noted, “My boyfriend and I attended a wedding, and stayed at my friend’s home. That night, we intended to engage in protected sexual behavior, but we couldn’t find any condoms, so we went without. ” (P9)

“I was on my first date with my girlfriend, we stayed with each other too late, so we didn’t go back to school because the dorm was closed. We ordered a hotel, but we didn’t prepare (condoms), and also the hotel didn’t have it (condoms) yet.” (P2)

Subtheme: personal use of money

Affordability shaped some students’ choices on condom purchases. They often prioritized spending on leisure activities over buying condoms, confident in their financial ability to handle any negative consequences of risky sexual behavior. “My family conditions are preferably, I think I have sufficient funds to deal with consequences, such as not wearing condoms or unintended pregnancy.” (P17).

“When my boyfriend and I went to the hotel, although the hotel had condoms, they were expensive for us and we didn’t buy and use them.” (P8)

Subtheme: improper use of alcohol

Alcohol consumption was commonly reported before engaging in risky sexual behavior. Some participants reported that drinking often lowered their inhibitions and impaired their judgment, leading to decisions that increased their risk of unprotected sex. “I went out to have dinner with some friends and a boy tried to pursue me, I drank a little too much beer at night. Then he took me to a hotel and we had sexual behavior when I woke up found it was unprotected.”(P5)

Theme 2: external influencing

Subtheme: influencing of sexual partners

A partner’s appearance, occupation, attitude or behavior may lead to risky sexual decisions. “I met a girl while traveling, she was very beautiful and sweet; shortly after chatting, she invited me to her apartment where we had sex.” (P12)

Some students mentioned that the pressure to conform to a partner’s wishes was a significant factor in engaging in risky sexual behavior. They often felt embarrassed to refuse their partner’s request to not use condoms, leading to unprotected sexual encounters. “At that time, unprotected sexual behavior was initiated by my partner, he didn’t want to wear (condom), and I was embarrassed to refuse him.”(P7)

Subtheme: influencing of peer

Most students would communicate with peers about sexual topics, and while some engaged in risky sexual behavior in order to “fit in” a group. “My roommates would share some of their sexual experiences with me. Initially, I did not understand these. But as their communication became more frequent, I gradually became aware that if I didn’t do that, I couldn’t fit in with them.” (P4)

“Many of my friends had engaged in these behaviors (sexual behaviors), and I was influenced by their example—being around peers who engaged in these activities gradually normalized them for me, so I began to go downhill, and then began to talk about sexual behavior.” (P19)

Subtheme: impact of social media

Many students reported that a significant portion of their sex-related knowledge came from the internet or social media, however, the inaccurate or misleading sexual information on the Internet has led to misconceptions about sex among them. And also a small number of students reported they had watched or imitated the scenes of pornography found on social media before or during having sexual activity. “I learned some sexual knowledge from the Internet, such as the safety period [i.e., the fertile window in the menstrual cycle when pregnancy is considered less likely] or without ejaculation [i.e., withdrawal/coitus interruptus], unprotected measures is accepted which would not get pregnant. So I think it is not a problem to go without wearing a condom before ejaculation.”(P2) Nearly one-third of students reported using social media to find sexual partners in this study. “I used social media Apps to find dating partners, I had also used all the available platforms.”(P7) “I usually look for sexual partners through social software such as Tantan, Momo, Soul and Weibo, and would make an appointment if someone also had ideas of having sexual behavior.”(P9)

Theme 3: emotional or physical status

Subtheme: negative emotion

Some students reported being driven to engage in risky sexual behavior due to feelings of loneliness, depression or stress. “I was afraid of loneliness and wanted someone to accompany me, but others might be willing to achieve a certain purpose (risky sexual behavior).”(P7) “The stress of studying really depresses me, I look for sexual encounters as a way to relax after the final examination.”(P11)

Subtheme: sexual needs or impulse

Sexual need is an inherent attribute of human beings, and it is also a fundamental physiological necessity. Some students said they had engaged in risky sexual behavior to satisfy their own sexual needs or desires. “I’m an adult, so naturally, I have desires and thoughts for this thing (sexual behavior). When my partner initiated it, I didn’t resist, and it just happened.” (P7) Additionally, some students mentioned acting impulsively leading to risky sexual behavior. “My partner and I did not use a condom because we acted on impulse and couldn’t control or consider the consequences in that moment.”(P2)

Theme 4: achieve a purpose

Subtheme: satisfied curiosity

Students engaged in risky sexual behavior often driven by specific motives, especially for the curiosity. “I had risky sexual behavior because of curiosity about sex between men / MSM sexual experiences, and sexual partner’s physical conditions (body, sexual organs).”(P11)

Subtheme: seek excitement

Some students engaged in risky sexual behavior to seek excitement to seek excitement and happiness, often motivated by the novelty of the experience. “I continued to date because there was a sense of novelty with sexual partner, when the novelty is gone and I would seek new sexual partner.”(P9) Additionally, some students would like to pay for sexual behaviors so as to satisfy the excitement. “I hadn’t tried paid sexual behavior before, nor had I tried it between two male and one female, and I thought it was exciting and I wanted to experience it, so why not try it?” (P14)

Theme 5: a specific opportunity

Subtheme: have mutual affection or trust

Some students reported that when he/she had good feelings, they could have sexual behavior or even risky sexual behavior. “While walking in the playground one evening, I encountered a girl. After chatting and feeling good connection, we decided to go to the hotel together. Believing that ejaculation would still take some time, therefore I did not use condom.”(P17)

Subtheme: lovelorn

Some students said that during or after a breakup, they would like to choose having risky sexual behavior in order to distract themselves from their previous relationship. “I couldn’t handle breaking up with my boyfriend, so I sought distraction by constantly engaging in sexual activity with strangers.”(P10)

Subtheme: unfamiliar environment

Very few students mentioned when they were in an unfamiliar environment, such as during traveling, increased their likelihood of engaging in risky sexual behaviors. “When I travelled to a certain city, I felt it might not matter if I did something risky in an unfamiliar place if I did something bad (risky sexual behavior) in a strange environment.”(P12)

Mixed-methods integration

To move beyond separate quantitative and qualitative reporting, a joint display was developed to integrate statistically significant predictors with qualitative themes and to generate mixed-method inferences regarding situational mechanisms of risky sexual behaviors (Table 5).

Table 5.

Mixed-methods integration of quantitative and qualitative findings

Integrated mechanism	Quantitative evidence	Qualitative explanation	Type of integration	Mixed-method inference
Material and substance-related risk contexts	Risky sexual behavior was reported by 37.3% of sexually active students. Alcohol-related sex and non-use of condoms were among the most common risky behaviors. Monthly living expenses and frequent entertainment venue visits were significantly associated with risky sexual behavior; visiting entertainment venues more than four times per month remained an independent risk factor.	Participants described condom unavailability, reluctance to buy expensive condoms, alcohol use before sex, and increased risk in hotels or entertainment settings.	Convergence and expansion	Material conditions shaped immediate sexual decision-making. Alcohol use, insufficient condom access, and entertainment-related contexts created proximal situations in which risky sexual behavior was more likely to occur.
Partner and peer influence	Frequent discussion of sexual topics with others was strongly associated with risky sexual behavior and remained an independent risk factor. Perceived social support also differed between risky and non-risky groups in univariate analysis.	Participants reported partner pressure not to use condoms, embarrassment in refusing partners, peer discussions of sexual experiences, and pressure to fit in with peers.	Convergence and expansion	Social relationships functioned as both risk and protection. Peer norms and partner pressure may normalize risky sexual behavior, whereas positive support may reduce emotional vulnerability and improve safer decision-making.
Digital and social media exposure	Daily social media usage was significantly associated with risky sexual behavior; spending 4–8 h per day on social media remained an independent risk factor.	Participants described obtaining inaccurate sexual information online, exposure to pornographic or sexually suggestive content, and using social media or dating apps to find sexual partners.	Convergence	Social media served as both an information environment and a partner-seeking environment, increasing exposure to misinformation, sexualized content, and opportunities for casual sexual encounters.
Sexual attitudes, emotion, and motivation	Agreement with premarital sex and higher scores on attitude toward sexuality were associated with higher odds of risky sexual behavior. Male students also had higher odds of risky sexual behavior.	Participants described sexual impulse, curiosity, novelty seeking, excitement seeking, loneliness, academic stress, depression, and breakup-related distress as triggers for risky sexual behavior.	Expansion	Risky sexual behavior was shaped not only by sexual attitudes but also by emotional and motivational states. Permissive or conflicted sexual attitudes, negative emotions, sexual impulse, and sensation seeking may jointly increase vulnerability.
Relationship and opportunity situations	Relationship status was significantly associated with risky sexual behavior in univariate analysis. Several risky behaviors, including casual sex, multiple sexual partners, commercial sex, and group sex, were reported among sexually active students.	Participants described condom non-use in trusted romantic relationships, risky sexual behavior after breakup, sexual encounters during travel, and risk-taking in unfamiliar environments.	Expansion	Opportunity-related contexts influenced whether risk occurred. Trust in romantic relationships may reduce perceived condom necessity, whereas breakup, solitude, travel, and unfamiliar settings may weaken behavioral restraint.
Protection self-efficacy and behavioral capability	Protection self-efficacy differed significantly between groups and remained an independent protective factor in logistic regression. Sexual limit-setting self-efficacy differed between groups in univariate analysis.	Participants reported failing to prepare condoms, being unable to obtain condoms, or feeling embarrassed to refuse unprotected sex requested by partners.	Convergence and explanation	Behavioral capability was a key protective mechanism. Students with stronger protection self-efficacy may be better able to prepare condoms, negotiate condom use, refuse unsafe sex, and avoid high-risk situations.
Family and sociodemographic background	Gender, mother’s education, mother’s occupation, monthly living expenses, and sexual orientation were associated with risky sexual behavior in univariate analysis. In logistic regression, male gender and mother’s occupation remained significant.	Qualitative data provided only indirect explanations, mainly through money use, sexual curiosity, partner-seeking, and social context. Maternal occupation was not directly discussed by participants.	Partial divergence	Some sociodemographic associations may reflect broader family, economic, cultural, or social background, but the qualitative findings only partially explained these patterns. These results should therefore be interpreted cautiously.

Overall, the integrated findings indicate that risky sexual behaviors among college students were shaped by both individual characteristics and immediate situational contexts. Quantitative findings identified demographic, behavioral, attitudinal, and psychological factors associated with risky sexual behavior, while qualitative findings explained how these factors operated through alcohol use, condom inaccessibility, partner and peer influence, social media exposure, emotional distress, sexual impulse, relationship dynamics, and opportunity-related contexts.

Discussion

This study provides a context-sensitive explanation of risky sexual behaviors among Chinese college students by integrating quantitative predictors with qualitative accounts of real-world sexual decision-making. Rather than focusing solely on stable individual-level factors, the findings show that risky sexual behaviors emerged from the interaction between psychosocial characteristics and proximal situational contexts, including alcohol use, condom accessibility, partner and peer influence, social media exposure, emotional distress, sexual impulse, relationship dynamics, and unfamiliar environments. This integrated perspective extends previous research by clarifying not only which factors are associated with risky sexual behaviors, but also how and under what circumstances these behaviors occur.

The results showed that 25.72% of college students reported having engaged in sexual activity. This finding is consistent with developmental theory suggesting that emerging adulthood, typically spanning the late teens through the twenties, is a period characterized by increased autonomy, identity exploration, and romantic or sexual exploration [28]. It is also broadly consistent with previous evidence on the occurrence of sexual behavior among college students in mainland China [29]. The overall prevalence of risky sexual behavior was 9.60%, rising to 37.3% among sexually active students, indicating that sexual initiation is a critical stratifying factor for risk exposure and intervention prioritization.

Alcohol-related sexual behavior was the most frequently reported risky behavior in this study, accounting for 24.06% of sexually active students. Alcohol consumption may impair cognitive judgment, reduce self-regulation, and increase susceptibility to unprotected sexual behavior. This finding is consistent with previous evidence showing an association between alcohol use and sexual risk-taking among college students and young people [15]. Event-level research also suggests that alcohol use before or during sexual encounters may be associated with reduced protective behaviors, particularly condom use, in specific sexual situations [16, 17]. Unprotected sexual intercourse remained common in the present study, indicating persistent gaps in condom use and the continued need for strengthened sexual health education, condom negotiation skills, and improved access to preventive resources [2, 10, 17].

Comparative analysis showed that the prevalence of risky sexual behavior in this study was comparable to findings from Hubei Province, where 7.2% of college students reported high-risk sexual behaviors [30], but lower than findings from Shanghai and southwest China, where the reported prevalence was 33.7% and 38.10%, respectively [31, 32]. These discrepancies may reflect regional variation in sociocultural norms, sexual attitudes, educational environments, and access to sexual health resources. Students in northeastern China may exhibit more conservative sexual attitudes shaped by traditional cultural values and limited comprehensive sexuality education. Methodological differences in sampling, measurement instruments, and operational definitions may also contribute to heterogeneity across studies. The substantially higher prevalence among sexually active students in the present study suggests that sexual initiation is closely linked to increased exposure to behavioral risk patterns, highlighting the need for targeted interventions that move beyond general education toward risk-stratified prevention strategies.

Theoretically, this study extends the explanatory scope of Social Cognitive Theory and the Information–Motivation–Behavioral Skills model by demonstrating that psychosocial determinants do not translate into sexual behaviors in isolation [12, 13]. Instead, their influence may be activated, weakened, or reshaped by proximal situational contexts. For example, protection self-efficacy may reduce risk only when students are able to access condoms and negotiate condom use with partners, whereas permissive sexual attitudes may increase risk when combined with alcohol use, emotional distress, sexual impulse, or partner-seeking opportunities. By mapping qualitative themes onto Belk’s situational elements, this study further provides a structured framework for understanding the material, social, antecedent-state, objective, and opportunity-related contexts in which risky sexual behaviors occur [25]. The mixed-methods integration revealed convergent determinants across both datasets, including economic status, family environment, entertainment venue exposure, social media use, relationship status, sexual attitudes, communication patterns, peer support, and sexual self-efficacy. Qualitative data further expanded these findings by highlighting sexual mental health, resource accessibility, and situational contexts as critical influences. This integration strengthens the explanatory value of the mixed-methods design by generating inferences across quantitative and qualitative findings [27].

To enhance conceptual clarity, the identified qualitative themes and subthemes were mapped onto Belk’s five situational elements [25] as follows: (1) material environment, including economic status, access to condoms, alcohol availability, and entertainment venues; (2) social environment, including partner characteristics, peer influences, and social media usage; (3) antecedent states, including attitudes toward premarital sex, sexual arousal or impulse, and negative emotions; (4) objective elements, including psychological needs such as curiosity and excitement, sexual mental health, and dating applications; and (5) opportunity elements, including relationship status, solitude, and unfamiliar environments.

Material environment

The study identified economic status and resource accessibility, including access to condoms, alcohol, and entertainment venues, as key aspects of the material environment influencing risky sexual behaviors. In this sample, higher disposable income may have facilitated access to entertainment venues, hotels, or partner-seeking contexts, thereby increasing exposure to situational risk. Qualitative findings further showed that limited access to condoms and other protective measures may increase the likelihood of unprotected sexual behavior, which is consistent with prior qualitative evidence showing that contraceptive decision-making among adolescents and young adults is influenced by access, perceived need, partner dynamics, and contextual barriers [33]. However, this study also found that even when condoms were available, some students remained reluctant to obtain or use them because of perceived inconvenience, embarrassment, concerns about quality, or partner-related considerations. Previous studies among college students have similarly shown that perceived barriers to condom use include reduced pleasure, perceived knowledge of a partner’s sexual history, reduced intimacy, and negative emotions during condom purchase or acquisition [34, 35].

To address these challenges, universities and public health agencies should improve the accessibility of condoms and other preventive resources. This may include installing free condom dispensers in campus health centers, restrooms, or other discreet campus locations, as well as offering anonymous online ordering services. These measures may help reduce access barriers and promote safer sexual practices among students.

Social environment

Partners and peers are critical components of the social environment influencing risky sexual behaviors. Characteristics of sexual partners, such as attractiveness, trust, intimacy, and condom negotiation, were identified in this study as influential factors. Previous experimental and behavioral evidence suggests that perceived partner attractiveness can influence condom use intentions [36], and safer sexual communication is strongly associated with condom use [37]. These findings highlight the role of partner-related judgments and communication patterns in sexual decision-making. Peer communication and peer norms may also shape students’ attitudes toward premarital sex, multiple partnerships, and condom use. Internet use also emerged as an important situational factor in the social environment. This study found that college students often used social media not only for sexual information but also for social interaction and partner seeking. Prior research has shown that mobile internet access, online sexual solicitation, and online partner seeking are associated with sexual risk behavior among adolescents [38]. Evidence from Chinese college students engaging in casual sexual behavior further suggests that online dating platforms may play an important role in partner seeking and unsafe sexual behavior. In a cross-sectional study in China, 57.98% of college students who engaged in casual sexual behavior reported finding casual sexual partners through the internet [39]. Other studies among Chinese college students also suggest that sexual attitudes, online information environments, and condom-use determinants are associated with risky sexual behaviors [40, 41]. In the present study, prolonged engagement with sexually suggestive content and online partner-seeking platforms may have reinforced more permissive sexual attitudes and normalized behaviors such as premarital sex and multiple sexual partnerships.

To address these influences, digital platforms and virtual simulation tools may be used to create risk-situation scenarios that enhance students’ sexual health knowledge, improve practical decision-making skills, and provide access to online support groups or professional consultation. Sexual health educators should also focus on enhancing communication skills among college students and their partners. Guidance on forming healthy friendships, resisting peer pressure, and avoiding sexual transactions should be integrated into interventions to foster responsible social and sexual behavior.

Antecedent states

This study found that conflicted sexual attitudes, prior risky sexual experiences, negative emotions, and sexual impulse were important antecedent states preceding risky sexual behaviors. The association between attitudes toward sexuality and risky sexual experiences suggests that conflicted or negative sexual attitudes may be linked to behavioral risk [21, 22]. Previous research has also shown that depressed mood and negative affect are associated with unprotected sex among young people [22, 42]. Similarly, interviews in the present study indicated that loneliness, academic stress, depression, and emotional distress after relationship loss could impair judgment and lead to risky sexual behaviors, including unprotected sex, casual encounters, and behaviors with potential consequences such as unintended pregnancy and STIs. The qualitative findings further suggest that sexual desire, impulsivity, curiosity, and sensation seeking are important factors driving risky sexual behavior, particularly among male participants. This interpretation is consistent with studies showing that sensation seeking and fear of missing out may be associated with high-risk sexual behaviors among college students and young people [19, 20, 32]. Rather than describing these behaviors as solely biologically or hormonally determined, it may be more accurate to interpret them as resulting from the interaction between sexual arousal, impulsivity, emotional state, perceived opportunity, and limited risk appraisal. These insights highlight the need for targeted education and interventions to help young adults manage emotional challenges, regulate impulses, and develop safer decision-making strategies in sexual contexts.

Given these findings, educational systems should integrate mental health education with sexual and reproductive health education. Such integration may help students recognize emotional triggers, manage stress and loneliness, and make informed decisions regarding sexual health.

Objective elements

Psychological needs and social media use were critical objective elements influencing risky sexual behaviors. College students, characterized by curiosity and a desire to explore sexual and relational experiences, may seek pleasurable, novel, or stimulating experiences, which can increase exposure to sexual risk contexts. Previous research in Shanghai suggested that fear of missing out was associated with high-risk sexual behaviors among college students [20]. Previous studies among college students have shown that sexual sensation seeking, sexual compulsivity, personality traits, alcohol use, and other substance use are associated with sexual risk-taking and HIV-related risk behaviors, supporting the role of motivational and self-regulatory factors in risky sexual behavior [43, 44]. This study found that a substantial proportion of students were active daily users of social media and related platforms, with some participants seeking sexual partners through these channels. This behavior aligns with previous research showing that young people may use mobile internet and social media for sexual information, online solicitation, and partner seeking [38, 39]. Social media platforms enable students to interact with both familiar contacts and strangers, often dedicating substantial time to dating and relationship building. While these platforms may facilitate social connections, they may also amplify opportunities for risky sexual behaviors.

Given these findings, healthcare professionals and educators should leverage digital platforms to promote safer online practices. By incorporating sexual health education, open communication, and guidance on responsible social media use, a safer and more supportive digital environment can be created for college students.

Opportunity elements

This study identified relationship status, solitude, travel, and unfamiliar environments as opportunity-related factors influencing risky sexual behaviors. Students in relationships characterized by high trust or love may perceive lower risk and may therefore be less likely to use condoms. This finding is consistent with research showing that love, trust, and relationship factors play important roles in adolescent sexual decision-making [44]. Studies among Chinese university students also suggest that relationship status and romantic involvement are associated with sexual behavior and risk-related attitudes [45, 46]. However, some evidence indicates that romantic relationships may also foster a sense of responsibility toward oneself and one’s partner, suggesting that relationship status can function as either a risk or protective context depending on perceived trust, communication, and responsibility [46].

Qualitative findings further revealed that some students engaged in risky sexual behaviors as a way to cope with the emotional aftermath of previous relationships. Unfamiliar environments, such as travel or being away from school, family, or familiar social networks, also emerged as risk contexts. Evidence from a cohort study among UK university students also suggests that foreign travel during holidays is associated with increased alcohol and cannabis use, casual sex, and new sexual partnerships, supporting the role of unfamiliar or travel-related environments as opportunity contexts for sexual risk-taking [47]. These findings underscore the importance of targeted interventions addressing opportunity elements, including relationship dynamics, breakup-related distress, travel-related sexual risk, and safer decision-making in unfamiliar environments.

This study has several limitations. First, the sample was drawn from college students in a single city in northeastern China (Daqing), which may limit the generalizability of the findings due to regional differences in socio-cultural and economic contexts. Second, the exclusion of students with severe mental health conditions reduces applicability to this potentially vulnerable population. Third, the reliance on self-reported data introduces possible recall bias and social desirability bias, which may have led to underreporting of sensitive behaviors. Although anonymity and confidentiality were emphasized to reduce these effects, they cannot be fully excluded. Selection bias may also exist, as participation in an online survey on sexual behavior could be influenced by students’ attitudes and willingness to disclose sensitive information. In addition, the qualitative component used a relatively small purposively selected sample, which may limit transferability, although this approach is appropriate for exploratory qualitative research focused on depth rather than statistical generalization. Although classes were randomly selected within participating universities, the universities themselves were selected according to accessibility and type rather than randomly from all universities in China; therefore, selection bias may remain. Finally, the cross-sectional design of the quantitative phase limits causal inference, allowing only associations to be interpreted rather than causal relationships. Future studies should use larger and more diverse samples across multiple regions and adopt longitudinal designs to improve generalizability and better assess temporal relationships.

Conclusions

This sequential explanatory mixed-methods study showed that risky sexual behaviors among Chinese college students were shaped by both individual-level factors and proximal situational contexts. Quantitative results identified demographic, behavioral, attitudinal, and self-efficacy-related factors associated with risky sexual behaviors, while qualitative findings explained how these factors operated through alcohol use, condom inaccessibility, partner and peer influence, social media exposure, negative emotions, sexual impulse, relationship dynamics, and unfamiliar environments. These findings suggest that prevention efforts should move beyond general sexual health education and target the real-life situations in which risky sexual behaviors occur, particularly alcohol-related contexts, condom-use barriers, online partner-seeking, partner negotiation, and emotional distress. Future studies should further validate these situational mechanisms using larger, multi-regional, and longitudinal designs.

Authors’ contributions

HZ, TP: designed and supervised the research. SYS, RZ, JWY: interviewing and writing original draft. QSL, BBW and XHL: investigation. HZ and TP: reviewing and editing the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This study was funded by the National Social Science Foundation of China, Grant/Award Number: 22BRK009. Funding bodies had no influence on study design, data collection, data analysis, data interpretation or writing the manuscript.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This study followed the guidelines for human studies and was carried out based on the World Medical Association Declaration of Helsinki. Ethical approval was obtained from the Ethics Committee of Harbin Medical University of Daqing (No.HMUDQ 20220820111). All participants agreed to take part in the survey by giving electronic consent, and those involved in the qualitative research signed written informed consent. Additional safeguards were applied during the qualitative interviews because of the sensitive nature of the topic, including voluntary participation, the right to skip questions or withdraw, privacy protection, anonymization of interview data, and referral information for psychological support when needed.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.