Disordered eating behaviour is not associated with sexual risk taking behaviour amongst emerging adults attending a tertiary education institution in Coastal Kenya: a latent class analysis approach

doi:10.21203/rs.3.rs-2578325/v1

Background Sexual risk-taking behavior (SRTB) is a well-documented pathway to HIV acquisition in emerging adults (EmA) and remains common amongst African EmA. We aimed to describe the relationship between disordered eating behavior (DEB) and SRTB amongst EmA attending a tertiary educational institution at the Kenyan Coast. Methods We applied a cross-sectional design nested in a young adults’ cohort study. Eligibility included sexually active EmA aged 18-24 years. Three DEBs (emotional, restrained and external eating) were assessed using the Dutch Eating Behavior Questionnaire and analyzed using exploratory factor analysis. Seven SRTB indicators were assessed: i) non-condom use ii) casual sex iii) multiple sex partners iv) transactional sex v) group sex vi) age-disparate relationship and vii) anal sex, and grouped into low vs. high SRTB using latent class analysis. Logistic regression was used to assess the association between DEB and SRTB. Results Of 273 eligible participants (female, n =110 [40.3%]), the mean [SD] of emotional, restrained and external eating was 1.9 [0.6], 2.0 [0.6] and 3.0 [0.5] respectively. Overall, 57 (20.9%) were grouped into the latent high SRTB class. Emotional (Adjusted odds ratio {AOR [95% confidence interval, CI]: 1.0 [0.9 – 1.0], p = 0.398), restrained (AOR, 1.0 [CI: 0.9 – 1.1], p = 0.301) and External (AOR, 1.0 [CI: 0.8 – 1.2], p = 0.523) eating were not independently associated with latent high SRTB. Conclusion There was no significant association between DEB and SRTB. More studies in different African settings are needed to validate our findings in order to lay a strong evidence base for public health interventions on SRTB in this and similar settings.

Sexual risk-taking behaviour

disordered eating behaviour

emerging adults

latent class analysis

Africa

Sexual risk-taking behavior (SRTB) is a well-documented pathway to HIV acquisition in emerging adults (EmA) [1–7]. Evidence suggests that SRTB is common amongst EmA in sub Saharan Africa (sSA) [8] with studies reporting non-condom use ranging from 4.0% (non-condom use in the past 6 months) in South Africa [9] to 94.5% (non-condom use in past 12 months) in Madagascar [10]. It is not surprising that new HIV infections amongst young people including EmA contributed 31% of new HIV infections in sSA in 2019 [11]. Due to their developing brains [12, 13], EmA have high propensity to impulsivity which has been shown to increase the risk of SRTB [12–14]. Further, EmA in tertiary institutions of learning enjoy freedom from close parental monitoring which may potentially encourage SRTB [15].

SRTB in EmA in sSA has been well studied. Previous studies have explored several underlying risks including socio-demographic factors e.g. age [7, 16], education level [17, 18] and socio-economic status [7, 19], relationship/behavioral factors e.g. alcohol/substance abuse [16, 17], intimate partner violence (IPV)/sexual violence [20, 21] and number of sexual partners [22], knowledge/attitude/belief factors e.g. knowledge of HIV status [23], HIV knowledge [24] and perceived HIV risk [24], family/community factors e.g. social network affiliation [25–27], and mental/physical health factors e.g. sexually transmitted infections (STIs) [28], post-traumatic stress disorder (PTSD) [29] and depression [16, 29].

Neglected, yet important especially in sSA where there is increased urbanization, which is associated with lifestyle changes, is disordered eating behavior (DEB) and body image concerns. DEB like emotional eating, restrained eating and external eating have been associated with SRTB in high income countries (HICs) [30]. Weight gain or loss occasioned by DEB is associated with body image concerns [31]. On the other hand, people who are dissatisfied with their physical appearance report being less comfortable and confident in sexual interactions [31]. Insecurity or anxiety about relationships and sexual interactions negatively impacts condom use negotiation [32]. Moreover, sociocultural factors may impact DEB and eventually SRTB [33].

There is documented evidence of an association between DEB and SRTB among EmA in HICs [30]. However, much less is known about DEB and whether it is associated with SRTB among EmA in sSA. By focusing on DEB, the current study seeks to fill this knowledge gap. Specifically, we aim to describe the relationship between three DEB constructs (emotional, restrictive and external eating) and SRTB amongst EmA attending a tertiary education institution in Coastal Kenya.

Study design

We applied a cross sectional study design nested within a young adults’ cohort study (YACOS). In brief, YACOS is a longitudinal study that aimed to characterize sexually transmitted infections (STIs) amongst EmA at Pwani University (PU), a tertiary learning institution located in Kilifi County, Coastal Kenya. PU students aged 18–24 years and willing to undergo STI/HIV screening were eligible. Students who reported to know their HIV-1 positive infection status were excluded since they already had the primary end point (HIV). Eligible volunteers were enrolled and followed up three-monthly. For the nested sub-study, eligibility included sexually active participants (reported having had sexual intercourse in the last 3 months). Recruitment was done consecutively between July 29th and December 21st 2021.

Study Procedures

During recruitment, the study counselor gave a brief overview of the study including objectives and procedures. Thereafter, taking one student at a time, the study counselor first allocated a study ID and screened them for eligibility. Next, eligible students were taken through the informed consent process. Consenting participants were taken through the first set of assessments including anthropometric indicators (weight, height, and waist and hip circumferences) by the study counselor. The study counselor then handed over participants to the study clinician for the second set of assessments which included, first, a repeat of the anthropometrics for validation, eating behaviour, mental health, SRTB, substance use, and clinical assessments. Mental health, substance use and SRTB assessments were conducted through an audio-computer assisted self-interview (ACASI) [34]. Clinical assessment included assessment of self-reported STI symptoms. Participants reporting STI symptoms were treated as per national guidelines [35]. The study clinician then handed over the participants back to the study counselor for HIV self-testing. In brief, the study counselor first provided pretest counselling followed by a demonstration on how to use the self- test kit. Participants were then allowed to test themselves and show the results to the study counselor for data capture. Post-test counseling was then offered. Participants who turned out to be HIV-infected were referred to the PU HIV voluntary counseling and testing centre for a confirmatory test in line with Kenya national guidelines [36]. All HIV tests were done using the OraQuick® Rapid HIV 1/2 antibody test as per national guidelines [36]. Each participant received a reimbursement to compensate for their time. Finally, the study counselor gave them the return appointment.

Ethical considerations

Ethical approval was obtained from the KEMRI Scientific and Ethics Research Unit reference number KEMRI/SERU/CGMR-C/166/3925 and PU Ethics Review Committee (PUERC [Ref: ERC/PhD/003/2019]). Further, administrative approvals were obtained from National Council for Science and Technology (Ref: NACOSTI/P/19/1142) and PU (Ref: PU/DVCRE/RSCH/VOL.1/37).

Tools of data collection

Sexual risk-taking behavior (SRTB)

SRTB was assessed using a customized hybrid of the Tambua Mapema Plus (TMP) trial Risk assessment tool [37] and the Kilifi Health Risk Behavior Questionnaire (KRIBE-Q) [38]. In brief, the TMP tool focuses on sexual risk assessment, including transactional sex, multiple sex partners and group sex behavior. The KRIBE-Q has previously been validated and adopted for use amongst adolescents and young adults in Kilifi and proved to be a reliable tool (Gwet ACI = 0.82) [38]. Seven SRTB indicators were assessed; i) non-condom use (defined as not having used a condom at last sex in the last 3 months, ii) casual sex (defined as sex with a short term partner [where short term means a partner one has sex with for the first time and they do not intend to have any further sexual contact with them] in the past 3 months), iii) multiple sex partners (defined as having had more than one sex partner in the last 3 months), iv) transactional sex (defined as having been paid and/or paid someone money, bought gifts, bought alcohol or supported their living expenses for sex in the last 3 months), ( v) group sex (having had sex with more than one person at the same time in past 3 months), vi) age-disparate relationship (having a sex partner who is more than 10 years older in past 3 months), and vii) anal sex (defined as having had anal sex in past 3 months).

Disordered eating behaviour (DEB)

Emotional, restrictive and external eating were measured using the Dutch eating behaviour scale (DEBQ) [39]. The DEBQ was designed to measure eating styles that may attenuate or contribute to the development of overweight and obesity. In brief, the DEBQ is a 33-item, self-assessment scale for assessing three eating behavior domains: the restrained subscale (10 items [e.g., difficult to resist delicious foods]), the emotional eating subscale (13 items [e.g. desire to eat when irritated]), and the external eating subscale (10 items [e.g. eating when you feel lonely]). The emotional eating pattern corresponds to the tendency towards overeating in response to negative emotions. Restrained eating refers to restriction of food intake in order to prevent weight gain or promote weight loss. External eating means eating more in response to external food cues such as food’s sight, smell and taste. Participants were required to rate each item on a 5-point Likert scale ranging from 1 (seldom/never) to 5 (very often). Item scores for each subscale were added to obtain an overall subscale score; the overall subscale score was then divided by the number of subscale items to calculate score per subscale. Higher scores indicate a greater tendency to exhibit the subscale behavior.

Binge eating, defined as eating an amount of food that others would regard as unusually large, was assessed by an additional question on the DEBQ scale.

Anthropometric indicators

Anthropometrics were measured in line with WHO protocols [40]. All measurements were independently taken by two assessors to the nearest 0.1 unit. Weight was measured using a digital scale (Model 769, Seca, Hamburg, Germany) and height was measured using a portable adult height scale (Model 37–113, HEMC, Noida, India). Waist and hip circumference were measured using a measuring tape. The mean of the two sets of measurements from the two assessors was used to compute Body Mass Index (BMI) [defined as weight in kilograms divided by the square of height in meters]) and Waist-Hip-Ratio (WHR). The mean BMI from assessor 1 and assessor 2 were not statistically different [t = -1.9610; df = 272; p = 0.051]. However, the mean WHR from assessor 1 was statistically different from assessor 2 [t = 4.6784; df = 272; p = 001]. A BMI of < 18.5 kg/m², 18.5–25 kg/m² and > 25 kg/m² was categorized as low, normal and high respectively [41]. A WHR of < 0.95 or < 0.8 for males and females respectively was classified as low risk for cardiovascular complications [40].

Mental health disorders

Binge drinking, and substance use were assessed. Binge drinking, defined as having six or more drinks on one occasion, was assessed using the Alcohol Use Disorders Identification Test (AUDIT) [42]. The AUDIT is a 10-question alcohol assessment tool developed by the WHO to assess alcohol consumption, drinking behaviors, and alcohol-related problems [42]. The screening tool was designed to be used globally and has been proven by various studies to be reliable and valid in detecting alcoholic tendencies [42–44]. Each AUDIT response is scored based on a point scale from 0 to 4. A score of 0, 1–7, 8–14 and above 14 represents abstainer, low risk consumption, harmful consumption and alcohol dependence respectively [42].

Other measures

These were administered through interviews and ACASI and included self-reported STI symptoms in the last three months, socio-demographic characteristics including age, sex and religion, HIV risk perception, gambling and injuries. Based on the Kenyan national STI guidelines, the presence of at least one of the following 10 symptoms in the last three months was regarded as having STI: urethral discharge, genital sores, dysuria (pain while urinating), testicular pain/tenderness, vaginal itching/burning, vaginal discharge, lower abdominal pain, pain while having sex, rectal discharge/bleeding and peri-anal sores/growths [35, 45].

Analysis

Distribution of study participants by socio-demographic, clinical, nutritional, DEB, and mental health indicators was assessed. Continuous data were presented using means and standard deviations while categorical data were presented using frequencies and percentages. The analysis then proceeded as follows:

In a first step, we ran a latent class analysis (LCA) [46] to identify SRTB latent classes based on the seven SRTB indicators. The model with lowest Akaike information criterion (AIC) and Bayesian information criterion (BIC) values was regarded to be the best. Assignment of study participants to their respective SRTB latent classes was based on their posterior probabilities of class membership. Secondly, we used percentages to summarize the loadings of the SRTB indicators in each latent SRTB class in the optimal LCA model. Thirdly, we ran an exploratory factor analysis (EFA)[47] with the orthogonal varimax rotation to test the factorial structure of the DEBQ items. We used the following criteria to determine the final number of factors: i) an eigenvalue of ≥ 1.0, ii) a scree plot test, and iii) factor loadings of at least 0.50. We used the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett’s sphericity test to confirm the data’s factorability. Next, we ran a confirmatory factor analysis (CFA) [48] to assess the fit of the factorial structure identified in the EFA. Good model fit parameters were i) a Tucker-Lewis index (TLI) ≥ 0.95, ii) comparative fit index (CFI) ≥ 0.95, iii) root mean square error of approximation (RMSEA) < 0.06 and iv) the smallest root mean square residual (RMR) possible.

Finally, we conducted logistic regression to assess the association between DEB (emotional, restrained and external eating) and SRTB latent classes controlling for socio-demographic characteristics, anthropometric indicators, mental health disorders, HIV risk perception, gambling, injuries and substance use. For a start, univariable analysis was carried out and exposure variables with a p-value of < 0.2 were carried forward to the multivariable model. Analyses were done using Stata version 15.0 [49]. A p value of less than 5% was considered statistically significant.

Characteristics of study participants

Overall, 273 participants were eligible for the sub-study. The mean age, BMI and WHR was 21.0 [SD, 1.6] years, 21.5 [SD, 2.7] kg/m² and 0.8 [SD, < 0.1] respectively. Overall, majority of the participants were male (163 [59.7%]), aged 21–24 years old (160 [58.6%]), had never been tested for HIV infection (212 [77.7%]) and did not report any STI symptoms in the past 3 months (172 [63.0%]) (Table 1).

Table 1: Characteristics of sexually active emerging adults attending a tertiary learning institution in Coastal Kenya

Characteristics	Category	Male	Female	Overall
		(n = 163)	(n = 110)	(N=273)
		N [%]	N [%]	N [%]
*Socio-demographic and clinical indicators*
Age [years] Mean [SD]	-	21.2 [1.5]	20.7 [1.5]	21.0 [1.6]
Age group [Years]	18 – 20	57 [35.0]	56 [50.9]	113 [41.4]
	21 - 24	106 [65.0]	54 [49.1]	160 [58.6]
Year of study	Year 1 & 2	92 [56.4]	63 [57.3]	155 [56.8]
	Year 3 & 4	71 [43.6]	47 [42.7]	118 [43.2]
Study program was your first choice	No	47 [28.8]	37 [33.6]	84 [30.8]
	Yes	116 [71.2]	73 [66.4]	189 [69.2]
Ever been tested for HIV	No	126 [77.3]	86 [78.2]	212 [77.7]
	Yes	37 [22.7]	24 [21.8]	61 [22.3]
Have children?	No	156 [95.7]	99 [90.0]	255 [93.4]
	Yes	7 [4.3]	11 [10.0]	18 [6.6]
Parents alive?	Both parents alive	128 [78.5]	91 [82.7]	219 [80.2]
	One or both parents dead	35 [21.5]	19 [17.3]	54 [19.8]
Living arrangement	In campus	17 [10.4]	17 [15.4]	34 [12.5]
	Outside campus	146 [89.6]	93 [84.6]	239 [87.5]
Religion	Catholic	48 [29.5]	25 [22.8]	73 [26.7]
	Protestant or other Christian	113 [69.3]	81 [73.6]	194 [71.1]
	Muslim	2 [1.2]	4 [3.6]	6 [2.2]
STI symptoms past 3 months	No	138 [84.7]	34 [30.9]	172 [63.0]
	Yes	25 [15.3]	76 [69.1]	101 [37.0]
Perceived chances of getting HIV	Small chance	135 [82.8]	95 [86.4]	230 [84.3]
	Great chance	28 [17.2]	15 [13.6]	43 [15.7]
Ever taken PEP and/or PreP	No	155 [95.1]	108 [98.2]	263 [96.3]
	Yes	8 [4.9]	2 [1.8]	10 [3.7]
Ever taken part in gambling	No	39 [23.9]	91 [82.7]	130 [47.6]
	Yes	124 [76.1]	19 [17.3]	143 [52.4]
Seriously injured past 3 months	No	139 [85.3]	99 [90.0]	238 [87.2]
	Yes	24 [14.7]	11 [10.0]	35 [12.8]
Binge eating ever	No	88 [54.0]	67 [60.9]	155 [56.8]
	Yes	75 [46.0]	43 [39.1]	118 [43.2]
*Anthropometric indicators*
Body Mass Index [Mean/SD]		21.1 [2.3]	22.1 [3.1]	21.5 [2.7]
Body Mass Index categories	Low [<18.5]	17 [10.4]	9 [8.2]	26 [9.5]
	Normal [18.5 – 25]	136 [83.4]	81 [73.6]	217 [79.5]
	High [ >25]	10 [6.1]	20 [18.2]	30 [11.0]
Waist-Hip Ratio (WHR [Mean/SD])		0.8 [0.0]	0.8 [0.0]	0.8 [0.0]
Waist-Hip Ratio categories	Low risk	163 [100.0]	53 [48.2]	216 [79.1]
	High risk	0 [0.0]	57 [51.8]	57 [20.9]
*Mental health indicators*
Binge drinking past 3 months	No alcohol last 3 months	70 [42.9]	76 [69.0]	146 [53.5]
	No	66 [40.5]	28 [25.5]	94 [34.4]
	Yes	27 [16.6]	6 [5.5]	33 [12.1]
Marijuana use past 3 months	Never used marijuana	105 [64.4]	95 [86.4]	200 [73.3]
	No	18 [11.1]	3 [2.7]	21 [7.7]
	Yes	40 [24.5]	12 [10.9]	52 [19.0]
Tobacco use past 3 months	Never used tobacco	122 [74.9]	102 [92.8]	224 [82.0]
	No	21 [12.9]	4 [3.6]	25 [9.2]
	Yes	20 [12.2]	4 [3.6]	24 [8.8]
Khat use past 3 months	Never used khat	122 [74.9]	102 [92.7]	224 [82.1]
	No	16 [9.8]	3 [2.7]	19 [6.9]
	Yes	25 [15.3]	5 [4.6]	30 [11.0]
Other drug use past 3 months	No	149 [91.4]	99 [90.0]	248 [90.8]
	Yes	14 [8.6]	11 [10.0]	25 [9.2]
*Disordered Eating behaviour*
Emotional eating [M/SD]	-	1.9 [0.6]	1.9 [0.6]	1.9 [0.6]
Restrained eating [M/SD]	-	2.0 [0.5]	2.0 [0.6]	2.0 [0.6]
External eating [M/SD]	-	3.0 [0.5]	3.1 [0.5]	3.0 [0.5]

*Used at least one of the following in the past three months: shisha, glue, heroin, cocaine, methamphetamine

STI: Sexually transmitted Infection

WHR [Low risk for cardiovascular complications]: <0.95 or <0.8 for males and females respectively

WHR [High risk for cardiovascular complications]: ≥ 0.95 or ≥ 0.8 for males and females respectively

PEP: Post-exposure prophylaxis

PreP: Pre-exposure prophylaxis

Please insert Table 1 here

Disordered eating behaviour constructs

The mean [SD] of the three DEB constructs based on all 33 DEBQ items was 1.9 [0.6], 2.0 [0.6] and 3.0 [0.5] for emotional, restrained and external eating respectively (Table 1). Based on EFA, the 33 DEBQ items produced 9 factors with eigen values of > 1.0. The total variance explained by the 9 factors was 58.5%. Out of all 33 items, 15 were not included in the final model due to factor loadings of < 0.50. Specifically, factors 4–6 had composite reliability of below 0.70 and were dropped. Factors 7–9 had only one item each with factor loadings of ≥ 0.50 hence composite reliability could not be checked and were therefore dropped. The remaining 3 factors (factor 1–3) each had a composite reliability of > 0.70 and were therefore included in the final model (Table 2). In factor 1, 11 items had loadings of ≥ 0.50. In factor 2 and 3, there were 5 and 2 items respectively (Table 2). With the exception of items excluded due to factor loadings lower than 0.50, all DEBQ items loaded into the 3 factors the same way as in the original DEBQ tool hence the names of factors generated remained unchanged (i.e. Factor 1 [Emotional eating {EmoEat} – 11 items]; Factor 2 [Restrained eating {ResEat} – 5 items]; Factor 3 [External eating {ExtEat} – 2 items]). The EFA therefore confirmed the original three-factor structure of the DEBQ scale. Through a CFA the fit of the final 18-item model identified in EFA with three latent variables (EmoEat – 11 items; ResEat – 5 items and ExtEat – 2 items) was examined. Psychometric parameters including discriminant validity, convergent validity and composite reliability (CR) were tested. The average variance extracted (AVE) which should be ≥ 0.5 for convergent validity to be assured was 0.671 for ExtEat but low for EmoEat and ResEat (0.401 and 0.462 respectively). However, with AVE values ≥ 0.4 and CR ≥ 0.6 (0.8770 and 0.8002 for EmoEat and ResEat respectively), convergent validity was still present [50]. The final 18 – item model demonstrated acceptable parameters (Table 3, Table 4 and Table 5).

Table 2

Loadings for all DEBQ items in the 9 factors retained after exploratory factor analysis
DEBQ Items^a	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	Factor 6	Factor 7	Factor 8	Factor 9
Item 1	0.638*	− 0.016	0.122	− 0.142	− 0.023	− 0.093	− 0.004	0.138	-0.139
Item 2	0.038	0.021	0.862*	0.013	0.001	− 0.011	− 0.057	− 0.067	0.082
Item 3	0.259	0.061	0.133	0.083	0.606*	− 0.324	0.163	− 0.085	− 0.162
Item 4	− 0.041	0.384	0.094	0.123	− 0.095	0.545*	− 0.224	− 0.154	− 0.099
Item 5	0.642*	0.083	0.259	− 0.045	− 0.156	− 0.067	0.177	− 0.079	0.059
Item 6	0.059	− 0.076	0.736*	0.173	0.245	0.196	0.022	0.054	0.002
Item 7	0.083	0.664*	− 0.023	− 0.071	− 0.187	0.087	0.028	0.037	− 0.026
Item 8	0.508*	− 0.022	0.147	− 0.043	0.143	0.154	0.229	− 0.034	0.142
Item 9	0.024	− 0.103	0.477	0.433	0.081	− 0.146	0.200	0.224	0.145
Item 10	0.687*	− 0.134	− 0.025	− 0.007	0.100	0.214	− 0.021	0.121	0.208
Item 11	0.035	0.351	0.007	− 0.024	0.526*	0.077	− 0.286	0.129	0.147
Item 12	0.001	− 0.022	0.388	0.317	− 0.076	− 0.052	0.273	0.357	− 0.106
Item 13	0.709*	0.072	0.070	0.025	0.142	− 0.111	0.056	0.088	0.181
Item 14	0.041	0.167	0.004	− 0.089	− 0.025	0.001	− 0.028	0.824*	− 0.035
Item 15	− 0.098	− 0.037	0.068	0.717*	0.052	0.161	0.044	− 0.021	− 0.009
Item 16	0.338	− 0.042	0.109	0.076	0.070	0.690*	0.219	0.067	− 0.039
Item 17	0.047	0.399	− 0.107	− 0.027	0.164	0.114	0.583*	− 0.042	0.020
Item 18	0.110	− 0.066	0.409	0.394	0.227	0.051	− 0.172	− 0.094	− 0.370
Item 19	0.092	0.362	0.058	− 0.129	0.214	0.294	0.022	− 0.105	0.329
Item 20	0.624*	0.029	0.101	− 0.026	0.049	0.155	0.346	− 0.003	− 0.255
Item 21	− 0.046	0.086	0.134	0.160	− 0.070	− 0.054	− 0.010	− 0.051	0.740*
Item 22	0.002	0.765*	0.061	− 0.084	− 0.124	− 0.035	0.017	0.045	0.077
Item 23	0.729*	0.009	− 0.079	0.068	0.172	0.185	0.061	0.025	0.081
Item 24	− 0.018	− 0.008	0.219	0.604*	0.137	0.220	− 0.093	− 0.306	− 0.036
Item 25	0.769*	0.041	0.097	− 0.064	− 0.018	− 0.037	− 0.104	− 0.058	− 0.049
Item 26	− 0.033	0.814*	− 0.006	− 0.012	0.099	0.030	0.085	0.068	0.057
Item 27	− 0.013	− 0.024	0.293	0.117	0.641*	0.231	0.166	− 0.064	− 0.058
Item 28	0.542*	− 0.006	0.007	0.126	0.068	0.049	0.542	− 0.035	0.046
Item 29	0.012	0.781*	− 0.058	0.033	0.186	− 0.056	0.065	0.019	− 0.055
Item 30	0.606*	− 0.012	− 0.071	0.009	0.116	0.152	− 0.118	0.063	− 0.072
Item 31	0.081	0.586*	− 0.160	0.080	0.026	0.106	− 0.182	0.385	0.088
Item 32	0.758*	0.052	− 0.129	0.061	− 0.117	0.023	− 0.001	− 0.044	− 0.260
Item 33	0.074	0.002	0.055	0.702*	− 0.029	− 0.136	0.037	0.017	0.205
Eigen value	5.54823	3.54233	3.05325	1.45832	1.29622	1.20231	1.15840	1.06005	1.00009
^aOriginal DEBQ item numbers: Restrained eating [Item 4,7,11,14,17,19,22,26,29,31]; emotional eating [Item 1,3,5,8,10,13,16,20,23,25,28,30,32] external eating [Item 2,6,9,12,15, 18, 21, 24,27,33]
*Loadings > 0.50

Table 3

Construct validity and reliability of the final DEBQ model (n = 273)
Factors		Parameters
	AVE	Square root of AVE	CR
Emotional eating	0.401	0.6332	0.8770
Restrained eating	0.462	0.6797	0.8002
External eating	0.671	0.8191	0.7374

AVE – Average Variance Extracted CR – Composite Reliability

Convergent validity- Present if AVE is ≥ 0.5; 0.4 acceptable if CR is above 0.6

Discriminant validity (Square-root of AVE)- Present if correlation between latent constructs is less than square-root of AVE

Construct validity indicates how close/related are the items measuring a latent construct

Discriminant validity indicates how different the items measuring a latent construct are from others

Table 4

Correlation between eating behavior constructs in final model (n = 273)
Latent variables	Emotional eating	Restrained eating	External eating
Emotional eating	1.0000	0.0533	0.1295
Restrained eating	0.0533	1.0000	-0.0636
External eating	0.1295	-0.0636	1.0000

Table 5

Fit statistics of final DEBQ model (n = 273)
	(TLI)	(CFI)	RMSEA	SRMR
Emotional eating	0.917	0.934	0.076	0.045
Restrained eating	0.963	0.982	0.074	0.030
External eating	1.000	1.000	0.000	0.000
TLI- Tucker-Lewis index
CFI- Comparative fit index
RMSEA-Root mean square error of approximation
RMR- Root mean square residual

SRTB latent classes

Overall, 219/273 (80.2%) of the participants reported at least one of the seven SRTB indicators. The most common SRTB indicator was non-condom use (n = 144 [52.7%]). Based on the LCA, the two-class solution was the best fitting model by AIC and BIC (Table 6). The two latent classes generated were class one (the low SRTB class) and class two (the high SRTB class). Compared to class 2, class 1 comprised the majority of participants (216/273[79.1%] vs 57/273[20.9%]) (Fig. 1b). Four out of the seven SRTB indicators were significantly higher in participants belonging to the latent high SRTB class compared to those in latent low SRTB class (Fig. 1b). Although group sex was reportedly more in the latent high SRTB class compared to the latent low SRTB class, (6.8% vs 2.7%), the difference was not statistically significant (p = 0.130) (Fig. 1b).

Table 6

Model fit statistics for one class and two class models in the latent class analysis
Number of classes	AIC	BIC
1	1570.119	1595.386
2	1536.481	1590.623
AIC: Akaike’s Information Criterion
BIC: Bayesian Information Criterion

Associations between DEB and SRTB class membership

Logistic regression was used to explore the associations between DEB represented by the three DEB constructs (EmoEat, ResEat and ExtEat) and latent high SRTB. DEB was not independently associated with latent high SRTB even after adjusting for socio-demographic, anthropometric and mental health indicators. Specifically, EmoEat (Adjusted odds ratio {AOR [95% confidence interval, CI]: 1.0 [0.9–1.0] p = 0.398), ResEat (1.0 [0.9–1.1] p = 0.301) and ExtEat (1.0 [0.8–1.2] p = 0.523) were not independently associated with latent high SRTB (Additional file 1). In separate validation analyses, we assessed the association between the three DEB constructs and each of the seven SRTB indicators included in the LCA. In all the seven models, we did not find an association between DEB and SRTB except ResEat which reportedly showed an association with anal sex (Adjusted odds ratio {AOR [95% confidence interval {CI}: 1.2 [1.0–1.4]] p = 0.027). However, only 10/273 (3.7% [Figure 1a) of participants had engaged in anal sex hence the suggested association between ResEat and anal sex needs to be interpreted with caution (S1 – S7 Tables).

Figure 1: Distribution of participants by SRTB indicators

Please insert additional file 1 here uploaded as additional file because it is too wide

Contrary to our hypothesis, our findings do not suggest an association between any of the three constructs of DEB (emotional, restrained and external eating) and SRTB. We had hypothesised that DEB would be associated with SRTB as reported in previous studies [30, 51–53]. We propose four explanations for the absence of association. First, most studies reporting associations between DEB and SRTB were conducted in HICs[30, 51–53]. It is possible that the socio-economic differences between HICs (where most of the studies reporting an association between DEB and SRTB were conducted) and low and middle income countries (LMICs [(where the current study was conducted) could explain the differences in the findings. Specifically, there is an increased emphasis on thinness in HICs. Such emphasis could impact body perception and eventually eating habits leading to DEB [54]. One potential explanation for the increased emphasis on thinness in HICs is social pressure. Specifically, social pressure regarding thinness and attractiveness, internalization of societal beauty ideals and body dissatisfaction [55] which have been shown to be high in HICs [56]. EmA especially young women, are exposed to societal messages about the physical characteristics associated with being beautiful, the importance of being thin, and how being attractive is seen as a characteristic of being successful in life, relationships and career [33]. Because of the discrepancy between the actual body outlook of EmA and what they consider to be the ideal body image, this is thought to lead to body image concerns and eventually DEB [57]. Apart from social pressure, another aspect of the sociocultural model explaining the increased emphasis on thinness in HICs is the influence of modern media which people in HICs have easy access to compared to LMICs [54]. Even with increasing urbanization, media access in LMICs remains comparatively lower. The disparity in media access, and consequently the influences of media, between the two settings imply a variation in societal preference for thinness and may explain the lack of association between DEB and SRTB in LMICs whereas an association has been found in HICs.

Secondly, previous studies assessed SRTB outcomes in isolation without considering that SRTBs tend to cluster together [30, 51–53]. Taking SRTB clustering into account, our study combines seven SRTBs into one endpoint. It is possible that the differences in outcome definition between our study and previous studies may have contributed to the observed differences in results. One study however did not find an association between DEB and SRTB just like ours [53]. The study was carried out in USA and enrolled adolescent women aged 14 years and above. The study found that women with DEB experienced later ages of first birth and lower parity compared to their counterparts without DEB.

Thirdly, we used the DEBQ tool to assess DEB. Since the tool has not been validated for use in the local Kenyan context, this may partially explain the lack of association with SRTB.

Finally, it is possible that there simply is no association between DEB and SRTB in this population.

Our findings imply that in LMICs like Kenya, interventions targeted at DEB among EmA towards controlling SRTB are unnecessary since our findings suggest lack of an association between the two. However, our results need to be interpreted with caution given that some studies have reported an association between DEB and SRTB as well as other risky behaviors including alcohol use [58], tobacco use [59] and illicit drug use. Further, DEB may adversely impact other health indicators including mental health disorders in developed settings [60], though this has not been extensively studied in LMICs and warrants further investigations.

We recommend more studies in Africa and other low-income settings to validate our findings that there is no association between DEB and SRTB. This is important especially now that Africa is experiencing an increased rate of urbanization. With increasing urbanization, body image concerns which potentially predispose to DEB are gaining momentum in Africa especially among urban EmA.

Our study is not without limitations. First, we collected data on binge eating and controlled for it in our analysis. However, we did not collect data on similar eating behaviors like anorexia and bulimia. This may have negatively impacted our ability to disentangle the relationship between DEB and SRTB which was not existent in our findings. Secondly, we used the DEBQ tool, which has not been validated for the local context, and may have resulted to the lack of an association with SRTB. There is need to develop and validate locally relevant eating behavior tools that can be applied and are appropriate for LMICs. Finally, the mean WHR obtained by assessor one differed from that of assessor two which may have impacted our analysis. However, none of the three DEB constructs showed an association with latent high SRTB even after excluding WHR from the multivariable logistic regression model (data not shown).

In conclusion, we obtained two latent SRTB classes from seven SRTBs and observed that there was no significant association between DEB and latent high SRTB. Our findings suggest that more studies, in different sSA settings are needed to find if indeed there is no association between DEB and SRTB in order to lay a strong evidence base for public health interventions on SRTB in this and similar settings.

Table 1

Characteristics of sexually active emerging adults attending a tertiary learning institution in Coastal Kenya
Characteristics	Category	Male	Female	Overall
		(n = 163)	(n = 110)	(N = 273)
		N [%]	N [%]	N [%]
Socio-demographic and clinical indicators
Age [years] Mean [SD]	-	21.2 [1.5]	20.7 [1.5]	21.0 [1.6]
Age group [Years]	18–20	57 [35.0]	56 [50.9]	113 [41.4]
	21–24	106 [65.0]	54 [49.1]	160 [58.6]
Year of study	Year 1 & 2	92 [56.4]	63 [57.3]	155 [56.8]
	Year 3 & 4	71 [43.6]	47 [42.7]	118 [43.2]
Study program was your first choice	No	47 [28.8]	37 [33.6]	84 [30.8]
	Yes	116 [71.2]	73 [66.4]	189 [69.2]
Ever been tested for HIV	No	126 [77.3]	86 [78.2]	212 [77.7]
	Yes	37 [22.7]	24 [21.8]	61 [22.3]
Have children?	No	156 [95.7]	99 [90.0]	255 [93.4]
	Yes	7 [4.3]	11 [10.0]	18 [6.6]
Parents alive?	Both parents alive	128 [78.5]	91 [82.7]	219 [80.2]
	One or both parents dead	35 [21.5]	19 [17.3]	54 [19.8]
Living arrangement	In campus	17 [10.4]	17 [15.4]	34 [12.5]
	Outside campus	146 [89.6]	93 [84.6]	239 [87.5]
Religion	Catholic	48 [29.5]	25 [22.8]	73 [26.7]
	Protestant or other Christian	113 [69.3]	81 [73.6]	194 [71.1]
	Muslim	2 [1.2]	4 [3.6]	6 [2.2]
STI symptoms past 3 months	No	138 [84.7]	34 [30.9]	172 [63.0]
	Yes	25 [15.3]	76 [69.1]	101 [37.0]
Perceived chances of getting HIV	Small chance	135 [82.8]	95 [86.4]	230 [84.3]
	Great chance	28 [17.2]	15 [13.6]	43 [15.7]
Ever taken PEP and/or PreP	No	155 [95.1]	108 [98.2]	263 [96.3]
	Yes	8 [4.9]	2 [1.8]	10 [3.7]
Ever taken part in gambling	No	39 [23.9]	91 [82.7]	130 [47.6]
	Yes	124 [76.1]	19 [17.3]	143 [52.4]
Seriously injured past 3 months	No	139 [85.3]	99 [90.0]	238 [87.2]
	Yes	24 [14.7]	11 [10.0]	35 [12.8]
Binge eating ever	No	88 [54.0]	67 [60.9]	155 [56.8]
	Yes	75 [46.0]	43 [39.1]	118 [43.2]
Anthropometric indicators
Body Mass Index [Mean/SD]		21.1 [2.3]	22.1 [3.1]	21.5 [2.7]
Body Mass Index categories	Low [< 18.5]	17 [10.4]	9 [8.2]	26 [9.5]
	Normal [18.5–25]	136 [83.4]	81 [73.6]	217 [79.5]
	High [> 25]	10 [6.1]	20 [18.2]	30 [11.0]
Waist-Hip Ratio (WHR [Mean/SD])		0.8 [0.0]	0.8 [0.0]	0.8 [0.0]
Waist-Hip Ratio categories	Low risk	163 [100.0]	53 [48.2]	216 [79.1]
	High risk	0 [0.0]	57 [51.8]	57 [20.9]
Mental health indicators
Binge drinking past 3 months	No alcohol last 3 months	70 [42.9]	76 [69.0]	146 [53.5]
	No	66 [40.5]	28 [25.5]	94 [34.4]
	Yes	27 [16.6]	6 [5.5]	33 [12.1]
Marijuana use past 3 months	Never used marijuana	105 [64.4]	95 [86.4]	200 [73.3]
	No	18 [11.1]	3 [2.7]	21 [7.7]
	Yes	40 [24.5]	12 [10.9]	52 [19.0]
Tobacco use past 3 months	Never used tobacco	122 [74.9]	102 [92.8]	224 [82.0]
	No	21 [12.9]	4 [3.6]	25 [9.2]
	Yes	20 [12.2]	4 [3.6]	24 [8.8]
Khat use past 3 months	Never used khat	122 [74.9]	102 [92.7]	224 [82.1]
	No	16 [9.8]	3 [2.7]	19 [6.9]
	Yes	25 [15.3]	5 [4.6]	30 [11.0]
Other drug use past 3 months	No	149 [91.4]	99 [90.0]	248 [90.8]
	Yes	14 [8.6]	11 [10.0]	25 [9.2]
Disordered Eating behaviour
Emotional eating [M/SD]	-	1.9 [0.6]	1.9 [0.6]	1.9 [0.6]
Restrained eating [M/SD]	-	2.0 [0.5]	2.0 [0.6]	2.0 [0.6]
External eating [M/SD]	-	3.0 [0.5]	3.1 [0.5]	3.0 [0.5]
*Used at least one of the following in the past three months: shisha, glue, heroin, cocaine, methamphetamine
STI: Sexually transmitted Infection

WHR [Low risk for cardiovascular complications]: <0.95 or < 0.8 for males and females respectively

WHR [High risk for cardiovascular complications]: ≥ 0.95 or ≥ 0.8 for males and females respectively

PEP: Post-exposure prophylaxis

PreP: Pre-exposure prophylaxis

SRTB – Sexual risk-taking behaviour

DEB – Disordered Eating Behaviour

EmA – Emerging Adults

sSA – Sub- Saharan Africa

PTSD- Post traumatic stress disorder

HIC – High income countries

LIC – low income countries

Ethical approval and consent to participate

Ethical approval was obtained from the KEMRI Scientific and Ethics Research Unit reference number KEMRI/SERU/CGMR-C/166/3925 and PU Ethics Review Committee (PUERC [Ref: ERC/PhD/003/2019]). Further, administrative approvals were obtained from National Council for Science and Technology (Ref: NACOSTI/P/19/1142) and PU (Ref: PU/DVCRE/RSCH/VOL.1/37).

Consent for publication

Not applicable

Availability of data and materials

The data sets used during the current study will be made available in a public repository

Competing interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.

Funding

This work was partially funded by the Medical Research Council [Grant number MR/M025454/1] to AA. This award is jointly funded by the UK Medical Research Council (MRC) and the UK Department for International Development (DFID) under MRC/DFID concordant agreement and is also part of the EDCTP2 program supported by the European Union. The authors are also grateful for the support of the Sub-Saharan African Network for TB/HIV-1 Research Excellence (SANTHE), a DELTAS Africa Initiative [grant number DEL-15–006] to AH. The DELTAS Africa Initiative is an independent funding scheme of the African Academy of Sciences (AAS)’s Alliance for Accelerating Excellence in Science in Africa (AESA) and supported by the New Partnership for Africa’s Development Planning and Coordinating Agency (NEPAD Agency) with funding from the Wellcome Trust [grant number 107752/Z/15/Z] and the UK government. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. This report was published with permission from the Kenya Medical research Institute (KEMRI).

Author contributions

All authors contributed to the study conception and design. SA added the eating behaviour concept to the study. EO, IM and SC collected data. Data management was done by CN. SC, SA, AH, K M and AK analysed the data. SC prepared the first draft manuscript. All authors revised and approved the final version for publication.

Price J, Pettifor A, Selin A, Wagner RG, MacPhail C, Agyei Y, et al. The association between perceived household educational support and HIV risk in young women in a rural South African community (HPTN 068): A cross sectional study. PLoS ONE. 2019;14(1):e0210632.
Price JT, Rosenberg NE, Vansia D, Phanga T, Bhushan NL MB, Brar SK HM, Tang JH, , et al. Predictors of HIV, HIV Risk Perception, and HIV Worry Among Adolescent Girls and Young Women in Lilongwe, Malawi. J Acquir Immune Defic Syndr. 2018;77(1):53-63.
Chimoyi LA and Musenge E. Spatial analysis of factors associated with HIV infection among young people in Uganda, 2011. 2014.
Mafigiri R, Matovu JK, Makumbi FE, Ndyanabo A, Nabukalu D, Sakor M, et al. HIV prevalence and uptake of HIV/AIDS services among youths (15-24 Years) in fishing and neighboring communities of Kasensero, Rakai District, South Western Uganda. BMC public health. 2017;17(1):251.
Schaefer R, Gregson S, Eaton JW, Mugurungi O, Rhead R, Takaruza A, et al. Age-disparate relationships and HIV incidence in adolescent girls and young women: evidence from Zimbabwe. AIDS (London, England). 2017.;31(10):1461-1470.
Mulu W, Yimer M and Abera B. Sexual behaviours and associated factors among students at Bahir Dar University: a cross sectional study. United Kingdom. 2014.
Obidoa CA, M'Lan CE and Schensul SL. Factors associated with HIV/AIDS sexual risk among young women aged 15-24 years in Nigeria. Journal of Public Health in Africa. 2012;3(1):59-64.
Chea SK, Kagonya VA, Abdullahi O, Abubakar AA, Abbeddou S, Michielsen K, et al. Sexual risk-taking behavior amongst emerging adults in Africa: a systematic review and meta-analysis. medRxiv. 2022;2022.09.13.22279893.
Andersson KM, Van Niekerk RM, Niccolai LM, Mlungwana ON, Holdsworth IM, Bogoshi M, et al. Sexual risk behaviour of the first cohort undergoing screening for enrolment into Phase l/II HIV vaccine trials in South Africa. International Journal of STD & AIDS. 2009;20(2):95-101.
Rahamefy OH, Rivard M, Ravaoarinoro M, Ranaivoharisoa L, Rasamindrakotroka AJ and Morisset R. Sexual behaviour and condom use among university students in Madagascar. Sahara J. 2008;5(1):28-35.
UNAIDS. UNAIDS DATA 2020. Geneva, Switzerland. 2020.
Ross JM, Coxe S Fau - Schuster RM, Schuster Rm Fau - Rojas A, Rojas A Fau - Gonzalez R and Gonzalez R. The moderating effects of cannabis use and decision making on the relationship between conduct disorder and risky sexual behavior. J Clin Exp Neuropsychol. 2015;37(1744-411X (Electronic)):303-15.
Ross JM, Duperrouzel J, Vega M and Gonzalez R. The Neuropsychology of Risky Sexual Behavior. J Int Neuropsychol Soc. 2016;22(1469-7661 (Electronic)):586-594.
Willen EJ, Cuadra A, Arheart KL, Post MJ and Govind V. Young adults perinatally infected with HIV perform more poorly on measures of executive functioning and motor speed than ethnically matched healthy controls. AIDS Care. 2017;29(1360-0451 (Electronic)):387-393.
Goldscheider FK and Davanzo J. Semiautonomy and leaving home in early adulthood. 1986.
Othieno CJ, Okoth R, Peltzer K, Pengpid S and Malla LO. Risky HIV sexual behaviour and depression among University of Nairobi students. Annals of General Psychiatry. 2015;14(1):16.
Nyembezi A, Sifunda S, Funani I, Ruiter R, Van Den Borne B and Reddy P. Correlates of risky sexual behaviors in recently traditionally circumcised men from initiation lodges in the Eastern Cape, South Africa. International Quarterly of Community Health Education. 2010;30(2):97-114.
Ososanya OO and Brieger WR. Rural-urban mobility in southwestern Nigeria: Implications for HIV/AIDS transmission from urban to rural communities. Health Education Research. 1994;9(4):507-518.
Pascoe SJS, Langhaug LF, Mavhu W, Hargreaves J, Jaffar S, Hayes R, et al. Poverty, food insufficiency and HIV infection and sexual behaviour among young rural Zimbabwean women. PLoS ONE. 2015;10(1):0115290.
Olley BO. Child sexual abuse as a risk factor for sexual risk behaviours among socially disadvantaged adolescents in Ibadan, Nigeria. Vulnerable Children and Youth Studies. 2008;3(3):243-248.
VanderEnde K, Chiang L, Mercy J, Shawa M, Hamela J, Maksud N, et al. Adverse Childhood Experiences and HIV Sexual Risk-Taking Behaviors Among Young Adults in Malawi. J Interpers Violence. 2018;33(11):1710-1730.
Yamanis TJ, Fisher JC, Moody JW and Kajula LJ. Young Men's Social Network Characteristics and Associations with Sexual Partnership Concurrency in Tanzania. AIDS and behavior. 2016;20(6):1244-1255.
Morris L, Kouya F, Kwalar R, Pilapil M, Saito K, Palmer N, et al. Factors associated with inconsistent condom use in adolescents with negative or unknown HIV status in Northwest Cameroon. AIDS Care. 2014;26(11):1440-5.
Ward-Peterson M, Fennie K, Baird S, Coxe S, Trepka MJ and Madhivanan P. Association between HIV awareness factors, health facility characteristics and risky sexual behaviour among young women in Zomba District, Malawi. Journal of biosocial science. 2018;50(6):853-867.
Mehra D, Kyagaba E, Ostergren PO and Agardh A. Association between self-reported academic performance and risky sexual behavior among Ugandan university students- a cross sectional study. Global journal of health science. 2014;6(4):183-195.
Mantsios A, Galai N, Mbwambo J, Likindikoki S, Shembilu C, Mwampashi A, et al. Community savings groups, financial security, and HIV risk among female sex workers in Iringa, Tanzania. AIDS and Behavior. 2018;22(11):3742-3750.
Karim AM, Magnani RJ, Morgan GT and Bond KC. Reproductive health risk and protective factors among unmarried youth in Ghana. International Family Planning Perspectives. 2003;29(1):14-24.
Masvawure TB, Terry PE, Adlis S and Mhloyi M. When ''no'' means ''yes'': the gender implications of HIV programming in a Zimbabwean university. J Int Assoc Physicians AIDS Care (Chic). 2009;8(5):291-8.
Peltzer K, Naidoo P, Matseke G, Louw J, McHunu G and Tutshana B. Prevalence of post-traumatic stress symptoms and associated factors in tuberculosis (TB), TB retreatment and/or TB-HIV co-infected primary public health-care patients in three districts in South Africa. Psychology, Health & Medicine. 2013;18(4):387-397.
Ramrakha S, Caspi A Fau - Dickson N, Dickson N Fau - Moffitt TE, Moffitt Te Fau - Paul C and Paul C. Psychiatric disorders and risky sexual behaviour in young adulthood: cross sectional study in birth cohort. BMJ . 2000;321(0959-8138 (Print)):263-6.
Cash T, Maikkula CL and Yamamiya Y. "Baring the body in the bedroom": Body image, sexual self-schemas, and sexual functioning among college women and men. Electronic Journal of Human Sexuality. 2004;7(.
Wingood GM, DiClemente Rj Fau - Harrington K, Harrington K Fau - Davies SL and Davies SL. Body image and African American females' sexual health. J Womens Health Gend Based Med. 2002;11(1524-6094 (Print)):433-9.
Striegel-Moore RH and Bulik CM. Risk factors for eating disorders. Am Psychol. 2007;62(3)(0003-066X (Print)):181-98..
Falb K, Tanner S, Asghar K, Souidi S, Mierzwa S, Assazenew A, et al. Implementation of Audio-Computer Assisted Self-Interview (ACASI) among adolescent girls in humanitarian settings: feasibility, acceptability, and lessons learned. Conflict and Health. 2017;10(1):32.
Ministry of Health Kenya and NASCOP. Kenya National guidelines for prevention, management and control of sexually transmitted infections Nairobi. 2018.
Ministry of Health Kenya and NASCOP. Guidelines for HIV testing services in Kenya. Nairobi. 2015.
Oxford Uo. Tambua Mapema Plus - to Discover HIV Infection Early and Prevent Onward Transmission (TMP). 2019.
Ssewanyana D, van Baar A, Newton CR and Abubakar A. A contextually relevant approach to assessing health risk behavior in a rural sub-Saharan Africa setting: the Kilifi health risk behavior questionnaire. BMC Public Health. 2018;18(1):774.
Tatjana van Strien JERF, Gerard P. A. Bergers ,Peter B. Defares. The Dutch Eating Behavior Questionnaire (DEBQ) for assessment of restrained, emotional, and external eating behavior. International journal of eating disorders. 1986;5(2):295-315.
WHO. Waist Circumference and Waist-Hip Ratio Report of a WHO Expert Consultation. Geneva. 2008.
WHO. Obesity : preventing and managing the global epidemic : report of a WHO consultation. Geneva. 2000.
Allen JP, Litten Rz Fau - Fertig JB, Fertig Jb Fau - Babor T and Babor T. A review of research on the Alcohol Use Disorders Identification Test (AUDIT). Alcohol Clin Exp Res. 1997;21(4)(0145-6008 (Print)):613-9.
Saunders JB, Aasland Og Fau - Amundsen A, Amundsen A Fau - Grant M and Grant M. Alcohol consumption and related problems among primary health care patients: WHO collaborative project on early detection of persons with harmful alcohol consumption--I. Addiction. 1993;88(3)(0965-2140 (Print)):349-62.
Saunders JB, Aasland Og Fau - Babor TF, Babor Tf Fau - de la Fuente JR, de la Fuente Jr Fau - Grant M and Grant M. Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on Early Detection of Persons with Harmful Alcohol Consumption--II. Addiction. 1993;88(6)(0965-2140 (Print)):791-804..
Abdul R, Gerritsen A, Mwangome M and Geubbels E. Prevalence of self-reported symptoms of sexually transmitted infections, knowledge and sexual behaviour among youth in semi-rural Tanzania in the period of adolescent friendly health services strategy implementation. BMC Infectious Diseases. 2018;18(.
Weller BE, Bowen NK and Faubert SJ. Latent Class Analysis: A Guide to Best Practice. Journal of Black Psychology. 2020;46(4):287-311.
Suhr DD. Principal component analysis vs. exploratory factor analysis. 2009.
DiStefano C and Hess B. Using Confirmatory Factor Analysis for Construct Validation: An Empirical Review. Journal of Psychoeducational Assessment. 2005;23(3):225-241.
StataCorp. Stata Statistical Software: Release 15. 2017.
Fornell C and Larcker DF. Evaluating Structural Equation Models with Unobservable Variables and Measurement Error. Journal of Marketing Research. 1981;18(1):39-50.
Baker JH, Johnson NK, Munn-Chernoff MA, Lichtenstein P, Larsson H, Maes HH, et al. Illicit Drug Use, Cigarette Smoking, and Eating Disorder Symptoms: Associations in an Adolescent Twin Sample. J Stud Alcohol Drugs. 2018;79(1938-4114 (Electronic)):720-724.
Fergus KB, Copp HL, Tabler JA-OX and Nagata JA-O. Eating disorders and disordered eating behaviors among women: Associations with sexual risk. Int J Eat Disord. 2019;52(1098-108X (Electronic)):1310-1315.
Tabler J, Schmitz RM, Geist C, Utz RL and Smith KR. Reproductive Outcomes Among Women with Eating Disorders or Disordered Eating Behavior: Does Methodological Approach Shape Research Findings? J Womens Health (Larchmt). 2018;27 (11)(1931-843X (Electronic)):1389-1399.
Pradeilles R, Holdsworth M, Olaitan O, Irache A, Osei-Kwasi HA, Ngandu CB, et al. Body size preferences for women and adolescent girls living in Africa: a mixed-methods systematic review. Public health nutrition. 2021;1-22.
Stice E. A prospective test of the dual-pathway model of bulimic pathology: mediating effects of dieting and negative affect. J Abnorm Psychol. 2001;110(0021-843X (Print)):124-35.
Rogers Wood NA and Petrie TA. Body dissatisfaction, ethnic identity, and disordered eating among African American women. J Couns Psychol. 2010;57(2):141-53.
Petrie TA, Tripp MM and Harvey P. Factorial and construct validity of the Body Parts Satisfaction Scale-Revised: An examination of minority and nonminority women. Psychology of Women Quarterly. 2002;26(3):213-221.
Fouladi F, Mitchell JE, Crosby RD, Engel SG, Crow S, Hill L, et al. Prevalence of Alcohol and Other Substance Use in Patients with Eating Disorders. Eur Eat Disord Rev. 2015;23(6)(1099-0968 (Electronic)):531-6.
Solmi M, Veronese N, Sergi G, Luchini C, Favaro A, Santonastaso P, et al. The association between smoking prevalence and eating disorders: a systematic review and meta-analysis. Addiction. 2016;111(11):1914-1922.
Tabler J and Geist C. Young Women with Eating Disorders or Disordered Eating Behaviors: Delinquency, Risky Sexual Behaviors, and Number of Children in Early Adulthood. Socius: Sociological Research for a Dynamic World. 2016;2(1-14.

No competing interests reported.

EBAdditionalfileBMCPH12022023.docx
Association between disordered eating behaviour and SRTB class membership (N=273)
EBSupportinginformationBMCPH12022023.docx

Disordered eating behaviour is not associated with sexual risk taking behaviour amongst emerging adults attending a tertiary education institution in Coastal Kenya: a latent class analysis approach

Status:

Version 1

Abstract

Figures

Background

Methods

Study design

Study Procedures

Ethical considerations

Tools of data collection

Sexual risk-taking behavior (SRTB)

Disordered eating behaviour (DEB)

Anthropometric indicators

Mental health disorders

Other measures

Analysis

Results

Characteristics of study participants

Disordered eating behaviour constructs

SRTB latent classes

Associations between DEB and SRTB class membership

Discussion

List Of Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1