Alcohol use, cigarette smoking, vaping and number of sexual partners: A cross‐sectional study of sexually active, ethnically diverse, inner city adolescents

Abstract Context There are few UK data on the prevalence and clustering of risky behaviours in ethnically diverse adolescents. Objectives To investigate the prevalence of reported alcohol use, smoking and vaping, and explore whether these behaviours are associated with increased numbers of sexual partners. Design Questionnaire survey of ‘Test n Treat’ chlamydia screening trial participants. Setting and participants Sexually active students attending six London technical colleges completed confidential questionnaires and provided genitourinary samples. Results The median age of the 509 participants was 17 years (IQR: 16‐18), 47% were male, 50% were of black ethnicity, 55% reported ≥2 sexual partners in the past year (67% of males and 45% of females) and 6.2% had chlamydia infection and 0.6% gonorrhoea. Almost half (48%) reported getting drunk in the past month, 33% smoked cigarettes and 7% had ever vaped. A larger percentage of students with ≥2 sexual partners than 0‐1 partners reported getting drunk in the past month (53.7%, 144/268% versus 42.2% 94/223, adjusted prevalence ratio: 1.33, 95% confidence interval: 1.11‐1.61) and smoking cigarettes (36.6%, 100/273% versus 30.2%, 67/222, 1.34 (1.05‐1.70)). By contrast, multiple sexual partners were not associated with vaping or chlamydia infection, but numbers were small. Conclusions We found high prevalences of risky behaviour and an association between multiple sexual partners and smoking and/or getting drunk. Findings support the introduction of compulsory sex and relationship education in UK secondary schools, including information about the adverse effects of alcohol and smoking. Public contribution Participants helped with study design, conduct and interpretation.


| BACKG ROU N D
There is on-going concern in many countries that increasing numbers of young adults undertake risky behaviours including getting drunk, cigarette smoking, vaping (using electronic cigarettes) and having multiple sexual partners. [1][2][3] In addition, many reports show that risky behaviours tend to cluster together. 1,2,4 A population-based survey of British young people aged 16-24 found participants reporting weekly or more frequent binge drinking (≥ weekly) or recent drug use (within the past 4 weeks) were more likely to report one or more new sexual partners in the past year. 5 Similarly in Italian adolescents, lifetime cannabis use, high alcohol intake and early sexual intercourse were more frequent among heavy smokers. 4 However, there are few UK data on the extent of clustering of risky behaviours, specifically in inner-city, ethnically diverse adolescents 6,7 ; there have been calls for more studies of substance use and sexual behaviour in teenagers. 2,5,8 Adolescence is a key stage in life when unhealthy habits are most often established, 9 and these can lead to health inequalities later in life. 1,10 Using cross-sectional, baseline data from the 'Test and treat' (TnT) feasibility trial (which explored the uptake of rapid chlamydia tests and same-day on-site treatment 11 ), we examined the prevalence of reported alcohol use, smoking (including heavy smoking defined as ≥10 cigarettes daily) and vaping, and explored possible associations with numbers of sexual partners (≥2 versus 0-1) within the past year.

| Participants
In September-October 2016, we recruited 509 sexually experienced male and female students aged 16-24 years. We approached students in public areas at six inner London further education/technical colleges 11 and asked whether they would consider helping us in a study on sexual health. Students who had never had sexual intercourse were ineligible. Following informed consent, students completed a confidential baseline questionnaire and supplied a self-taken genitourinary sample for later chlamydia/gonorrhoea testing using the Cobas 4800 CT/NG system (Roche Diagnostics), with positives confirmed using the Cepheid CT/NG GeneXpert system. Participants were given £5.
The questionnaire asked about date of birth, gender and ethnic origin. It also asked the following: Altogether in the last year, how many people have you had sexual intercourse with? One, Two, Three or more, None.
In the past month, how many times have you got drunk due to alcohol? 1-4, 5 or more, None. How many cigarettes do you smoke each day? More than 10, 1-10, None.

| Participant involvement
Four students on the trial steering committee gave advice throughout the study, 26 took part in qualitative interviews, 12 and findings were fed back to participants for their comments during an end of the study workshop.

| Statistical analysis
The sample size was determined to give adequate precision in the feasibility outcomes of the TnT study. 11 Statistical analyses were performed with STATA 10. The prevalence of getting drunk, smoking, vaping and multiple partners was presented as percentages with 95% confidence intervals (CIs). We used prevalence ratios (PRs) to explore possible associations between number of sexual partners (≥2 versus 0-1) and getting drunk, smoking or vaping at baseline.
Adjusted analyses were performed using multivariable Poisson models with a robust variance estimator, in order to model the prevalence ratio explicitly whilst stabilizing the variance. 13 Sensitivity analyses using log-binomial regression models were also undertaken and gave similar results.
Earlier studies suggested that the prevalence of risky behaviours may vary by ethnic group, 5,14 and higher rates of risky sexual behaviour in males than females. 5,7,15 Therefore, we adjusted multivariable analyses for two binary covariates: gender (male/female) and white ethnicity (yes/no).

| Sexual partners and risk behaviours
Of 499 responders, 55.1% (275, 95% CI: 50.8%-59.5%) reported that they had two or more sexual partners in the past year. Reporting multiple sexual partners was more common in males than females (66.5% vs 44.9%, prevalence ratio: 1.48 [95% CI: 1.26-1.74]) and in those of non-white rather than white ethnicity (58.0% vs 48.1%, 1.21 (0.99-1.47)). After adjusting for gender and ethnicity, reporting at least one episode of being drunk in the past month or being a regular cigarette smoker was more common in those reporting ≥2 sexual partners in the past year than those with 0-1 partners ( Table 2). This did not apply to vaping or heavy smoking, but numbers were small.

| Participant feedback
Student attenders at the end of the study workshop confirmed that most had received little education about risky sexual behaviours.
Sex and relationship education was 'mainly about putting on con-doms…….not to get pregnant'.

| Principal findings
Almost half of these young, multi-ethnic students reported getting drunk in the past month and a third were smokers, with both behaviours more common in those reporting multiple sexual partners.
Only 7% reported ever using e-cigarettes. There were no clear associations between risky behaviours and chlamydia infection, but numbers were small.

| Strengths and limitations
This is the largest UK community-based study of alcohol, smoking and sexual lifestyles in high-risk, inner-city adolescents. It included a high number of adolescents from ethnic minority groups, notably We only recruited students who said they were sexually active, and this may vary by culture. However, the association between male gender and risky sexual behaviour (such as having more sexual partners) is in line with other studies. 5,15 Other weaknesses are that we did not define what we meant by 'getting drunk' or 'sexual intercourse', nor did we ask about socio-economic status which may be associated with risky behaviour. 1 In addition, we could not look at students who had dropped out of classes at the colleges, or adolescents classified as 'NEETs' (Not in Education, Employment or Training). These could be young people with more risky behaviours.
Although we did not include those with special educational needs, we did include some participants with lower literacy levels who asked for help in completing the questionnaires. As vaping is a relatively new risk behaviour, we asked about ever vaping but used different timing for alcohol (drunk in past month) and smoking (cigarettes per day). It is likely that the rate of ever vaping in adolescents will increase over time. 2,16  However, a large 2010-12 study of British young people aged 18-24 found a slightly higher reported ever use of vaping: 12% versus 7% in our study. 16 Our participants were younger and more ethnically diverse, and 90% were adolescents. A Swedish study of 2185 teenagers

| Comparison with other studies
found reported e-cigarette use was associated with personal and parental tobacco use, physical inactivity and an unhealthy diet. 8 We did

| Implications for policy and practice
Findings could influence clinical practice and public health policy.
We found high prevalences of getting drunk, smoking and multiple sexual partners in sexually experienced, teenage further education college students in London with possible associations between getting drunk, cigarette smoking and increased number of sexual partners. These findings support calls for more education and health promotion for similar hard-to-reach, young or multi-ethnic populations. 5

| CON CLUS IONS
In the first UK community-based study of risky behaviours in sexually active, inner-city adolescents, we found high rates of reporting multiple partners (55%), getting drunk in the past month (48%) and cigarette smoking (33%). Participants with ≥2 partners in the past year were more likely than those with 0-1 partners to have got drunk in the past month and to smoke cigarettes. General practitioners doing health promotion in potentially vulnerable teenagers could consider asking about more than one risky behaviour. In addition, sex education for adolescents should be compulsory and address both sexual behaviours and substance use. 1,5,10 Policymakers should be aware of the urgent need for better health education for innercity teenagers.

ACK N OWLED G EM ENTS
We thank students and staff at the six participating further education/technical colleges.

AUTH O R S ' CO NTR I B UTI O N S
RB, PO, FR and SKB designed the study. AL did the chlamydia tests, and RB, SKB, N B-H, RP and FR analysed the data. RB wrote the first draft of the paper to which all authors including CF and FY then contributed. All authors read and approved the final manuscript.