Problematic online dating app use and its association with mental and sexual health outcomes in a sample of Men-having-Sex-with-Men

Abstract

Background and aims

Online dating applications (ODAs) are gaining popularity, raising concerns about their potential addictive effects on users' health. The main objective of this study was to investigate the association between problematic ODA use and mental health, substance use, and sexual behavior outcomes in men-who-have-sex-with-men (MSM). To achieve this, we first validated a German version of the Problematic ODA Use Scale (PODAUS), which assesses problematic ODA usage patterns.

Methods

Problematic ODA use was assessed in 226 HIV-negative MSM counseling for HIV pre-exposure prophylaxis at baseline and 6-month follow-up (n = 164, M_age = 42, SD = 11). The German PODAUS was validated using confirmatory factor analyses (CFA) and correlational analysis. Multiple linear regression analyses were performed to investigate associations of PODAUS with mental, sexual health, and substance use outcomes.

Results

CFA supported PODAUS’ 6-item unidimensional structure, demonstrating excellent fit indices. Reliability analyses indicated good internal consistency (α = 0.79) and high test-retest reliability (r_s = 0.68). Regarding mental health, significant positive associations were found between PODAUS and symptoms of depression, loneliness, anxiety, Attention Deficit Hyperactivity Disorder (ADHD), and impulsivity, with medium to large effect sizes. Regarding sexual health, significant negative associations of small to medium effect sizes were found between PODAUS and lifetime prevalence of sexually transmitted infections, body acceptance, and sexual fulfillment, while a significant positive association with trait sex motivation occurred.

Discussion and Conclusion

The German version of PODAUS emerges as a valid, reliable tool to assess problematic ODA use. In an MSM sample, problematic ODA use was associated with lower mental and sexual health.

Background

Given the omnipresence of smartphones and internet access, the extensive use of online dating apps (ODAs) necessitates an investigation into their potential associations with health outcomes. These apps, utilized by over 366 million users worldwide, are described as platforms for more than just dating, enabling romantic, sexual, and social connections (Anzani, Di Sarno, & Prunas, 2018; Curry, 2024). Grindr is the most widely used ODA among men-who-have-sex-with-men (MSM). With over 13.5 million users in 2023 and expected upward trends, ODA use is becoming increasingly prevalent (Statista, 2024). While ODAs are typically used by 40–50% of the general population, studies indicate a higher prevalence among MSM, with estimated usage rates reaching 78%, with many meeting sexual partners through these platforms and a significant number engaging daily (Badal, Stryker, DeLuca, & Purcell, 2018; Blanc, 2023; Sumter & Vandenbosch, 2019). Geolocation features in popular ODAs like Tinder and Grindr facilitate both online and offline interactions, leading to sexual encounters (Miles, 2017). ODAs offer quick, rewarding experiences through positive social feedback, fostering instant gratification and potentially addictive behaviors due to reward anticipation and dopaminergic activation before receiving interactions like messages or matches (Griffiths, 2018; Liberacka-Dwojak et al., 2023; Linnet, 2014). The unpredictability and intermittent reinforcement, combined with anonymity and local search capabilities, may reinforce continuous use and heighten the addictive potential of ODAs (Orosz, Tóth-Király, Bőthe, & Melher, 2016). Despite its advantages for users, ODAs may contribute to adverse mental health outcomes in MSM, such as low self-esteem, body dissatisfaction, and trauma from discrimination, harassment, and racism (Bineau et al., 2021; Frison & Eggermont, 2015).

This raises questions about the psychological, social, and health implications of prolonged and potentially addictive properties of these applications, coining the term “problematic ODA use” (Gori, Topino, & Griffiths, 2024; Orosz et al., 2016). A recent meta-analysis estimated the global prevalence of social media addiction at 14% and cybersex addiction at 8%, underscoring the relevance of investigating problematic ODA use, for which prevalence data are currently unknown (Meng et al., 2022).

Empirical evidence increasingly links extensive use of ODAs to mental health adversities in MSM, such as depression, anxiety, low self-esteem, diminished self-control, greater psychological distress, and suicidal thoughts (Holtzhausen et al., 2020; Lenton-Brym, Santiago, Fredborg, & Antony, 2021; Zlot, Goldstein, Cohen, & Weinstein, 2018). MSM, who are more likely to use ODAs, also face a 2.5-fold higher risk of mental disorders (Meyer, 2003), including depression, anxiety, suicide attempts, and addiction, compared to heterosexual men (Krüger, Pfister, Eder, & Mikolasek, 2022; Mongelli et al., 2019).

Studies on the link between ODA use and substance use show mixed results. While some studies show higher alcohol and cannabis use among ODA users, associations diminish when accounted for usage intensity (Flesia, Fietta, Foresta, & Monaro, 2021a; Flesia, Fietta, Foresta, & Monaro, 2021b). Chemsex, involving substances like methamphetamine, alkyl nitrites (poppers), cocaine, 3,4-methylenedioxymethamphetamine (MDMA), or γ-hydroxybutyric acid (GHB) during sexual encounters, is increasingly common among MSM ODA users (Choi et al., 2021; Herrijgers, Platteau, Vandebosch, Poels, & Florence, 2022).

Research has linked the ease of connecting with potential sexual partners via ODAs to increased sexual risks and risky sexual behaviors, such as unprotected intercourse, multiple partners, sex addiction, and a rise in sexually transmitted infections (STIs), although results are mixed (Wang et al., 2018; Winetrobe, Rice, Bauermeister, Petering, & Holloway, 2014). Moreover, higher levels of depression and anxiety among ODA users exacerbate risky sexual behavior, further impairing mental and overall health (Ciocca et al., 2022; Zlot et al., 2018).

To evaluate the associations of problematic ODA use with mental, sexual health, and substance use outcomes, a validated psychometric instrument is needed. Prior research has utilized Griffiths' biopsychosocial components model of addiction from 2005, which includes salience, mood modification, tolerance, withdrawal, conflict, and relapse, to develop various scales, like the Facebook Addiction Scale (FAS; Andreassen, Torsheim, Brunborg, & Pallesen, 2012) and the Problematic Tinder Usage Scale (PTUS; Orosz et al., 2016). However, most studies have focused on single-app usage, overlooking the complex nature of multi-app engagement (Coduto, Lee-Won, & Baek, 2020). Other instruments like the Online Dating Intensity Scale (Bloom & Dillman Taylor, 2020) and the Cyberdating Q_A (Sánchez, Muñoz-Fernández, & Ortega-Ruíz, 2015) target specific facets of online dating without a full psychometric evaluation, indicating a gap in measuring problematic ODA use comprehensively. However, the Problematic Online Dating App Use Scale (PODAUS; Gori et al., 2024) has been recently developed based on Griffiths' model and tested in an Italian convenience sample of young adults across multiple apps.

Despite significant insights, substantial gaps remain in understanding the prevalence and impact of problematic ODA use among MSM, including risk factors for its development and persistence. Conceptualizing and measuring problematic ODA use is still in its infancy, and profoundly exploring the health associations of problematic ODA use is missing. Additionally, PODAUS has only been tested in an Italian convenience sample, its suitability for a high-risk cohort for sexual health adversities, such as HIV-negative MSM counseling for HIV pre-exposure prophylaxis (PrEP), remains uncertain. Furthermore, the samples differ in age, linguistic, and cultural background, and no German version of the scale has been evaluated yet.

This study addresses these gaps through two primary objectives: first, to validate a German version of the PODAUS; and second, to examine the associations between problematic ODA use and mental health, substance use, and sexual health outcomes within a cohort of HIV-negative MSM counseling for HIV PrEP. Based on the aforementioned research, we hypothesized that higher intensity of problematic ODA use would be associated with lower mental health, higher substance use, and poorer sexual health outcomes.

Methods

Design

The present study, called Real Risks of the Online World, is a subproject of the ongoing Swiss-PrEPared study, a nationwide longitudinal observational cohort study and program enrolling HIV-negative individuals interested in taking PrEP (see swissPrEPared.ch and clinicaltrials.gov, identifier: NCT03893188). Participants receive regular PrEP counseling and STI screening, following the latest international recommendations (for detailed study design and cohort description, see Hovaguimian et al., 2022). Participants complete an online questionnaire on their smartphones seven days before their consultation. The questionnaire collects essential health information, including STI symptoms, and data on sexual behavior, including ODA use. The Real Risks of the Online World subproject, reported here, is a longitudinal study comprising a baseline assessment and a 6-month follow-up (see clinicaltrials.gov, identifier: NCT04714034). Data were collected between November 2021 and December 2022 at Checkpoint Zurich, Switzerland's largest sexual health clinic.

Participants

Participants were sought via advertisements, social media postings, and direct approaches during regular consultations at Checkpoint Zurich. Participants were eligible for participation if they (a) were aged ≥18 years; (b) active SwissPrEPared study participants at the Checkpoint Zurich; (c) used ODAs regularly in the past 12 months (ODAs were defined as all types of GPS-based social networking apps that run on a smartphone and are used to find potential relationships or sexual partners); and (d) sufficiently German-speaking. Potential participants were excluded if they were unable to follow study procedures. Figure 1 depicts the inclusion and exclusion process in a flowchart. The final study sample consisted of 226 participants at baseline and 164 at the 6-month follow-up assessment (see Table 1). The median age was 41.5 years (IQR: 32–50). 100% of participants were assigned male at birth, with 99% describing their current gender as male and 92% defined as homosexual. Additionally, 83% obtained a higher education, 57% reported living comfortably with their income, and 67% were born in Switzerland.

Fig. 1.

Flowchart displaying inclusion and exclusion criteria

Table 1.

Baseline sample characteristics (N = 226)

Age (in years)	Median [IQR]		41.5 [32–50]
Sex assigned at birth	Male		100.0%
Gender	Cis-malea		98.7%
	Trans-maleb		0.0%
	Non-binary		0.9%
	Not stated		0.4%
Sexual orientation	Homosexual		92.0%
	Bisexual		4.9%
	Heterosexual		0.9%
	Not defined		2.2%
Education	University/Higher Education		82.7%
	Apprenticeship		15.0%
	Mandatory Education (9 years)		1.3%
	No/other Education, no information		0.9%
Socioeconomic status	Living (really) comfortably		56.6%
	Neither un-/comfortable		13.3%
	(Really) Struggling		7.5%
	Did not state any information		22.6%
Birth country	Switzerland		68.6%
	Germany		14.6%
	Italy		3.1%
	Other countries		13.7%
Dating app preference	Grindr		92.0%
	PlanetRomeo		69.7%
	Instagram		36.7%
	Scruff		31.0%
	Hornet		2.2%
	Tinder		20.8%
	Other		19.5%
Got in contact with illegal substances via ODA (Either you or your sexual partner used substances on a date)			54.4%
Illegal substances were offered for sale on ODA			44.2%
Started taking illegal substances because of ODA use			12.0%
Used more illegal substances because of ODA			22.6%
		M	SD	Range
Nr of ODA used in past 12 months		2.7	1.1	1–6
Time spent on dating apps (in hours)
Weekly total		9.5	9.5	0–84
Weekday		1.3	1.5	0–15
Weekend day		1.6	1.5	0–15
Objective: Weekly Smartphone screentime (in hours)		5.1	8.8	0–98
Motivations for use of favorite ODA
Finding a sex partner		4.3	0.8	1–5
Getting sex fast		3.6	1.1	1–5
Chatting		3.3	1.2	1–5
Confirmation for looks		2.4	1.2	1–5
Finding a partner for life		2.3	1.3	1–5
Escaping my problems		2.1	1.2	1–5
Masturbating		1.9	1.2	1–5
Afraid of missing a potential partner		1.8	1.1	1.5

*Note. ^a Cis refers to individuals for whom sex assigned at birth matches gender identity, ^b Trans refers to a discrepancy between sex assigned at birth and the reported gender identity, ODA = Online Dating Apps, Other countries include Italy, Netherlands, Brazil, Spain, Austria, USA, Belgium, France, Greece, Kazakhstan, Mauritius, Philippines, Poland, Portugal, Paraguay, Romania, Russia, Sweden, Slovenia, Slovakia, Syria, Chad, South Africa.

Procedure

Participants completed self-reporting questionnaires through the REDCap electronic data capture tool, which took about 30 min (Harris et al., 2009). On-site interviews, conducted in German by trained psychologists, lasted approximately 90 min and included assessments of substance use and mental health, including professional evaluation using standardized interviews. Participants were compensated 45 USD for each study time point (baseline and follow-up).

Measures

Measures used in this study can be classified into four categories: 1) problematic ODA use, 2) mental health, 3) substance use, and 4) sexual health (see supplementary Table 2 for information on internal consistency and descriptive statistics at baseline of all included measures). Translation and back-translation processes were employed for scales unavailable in German (Brislin, 1970). Demographic information was sourced from the Swiss-PrEPared study, coinciding with the most recent prior study visit. Education was measured as a single item and coded into four categories.

Problematic ODA use

Problematic ODA use in the past 12 months was assessed with a modified German version of the 6-item Problematic Dating Apps Use Scale (PODAUS; Gori et al., 2024) see supplementary S1 for the German PODAUS items and Table S2 for an English translation of the German PODAUS. A distribution of response percentages for each PODAUS item is displayed in supplementary Table S3. A detailed description of item and scale development is outlined in supplementary materials S4. Participants were asked to indicate the ODAS they used, with the option to select multiple apps simultaneously. The listed apps were determined based on a literature review and focus group interviews in the Prepiloting. Additionally, participants could include other ODAs not listed via a free-text entry. Unlike the Italian version's 5-point Likert scale, the German adaptation used a 4-point scale, excluding a neutral option to minimize neutral response bias (Krosnick et al., 2001). Items were answered from 1 (do not agree at all) to 4 (agree completely), with higher scores indicating more problematic ODA use in the total sum score. Additionally, items related to mood modification and salience were adapted, with the German version focusing more on escapism and thoughts about app usage rather than direct usage. Self-reported weekly time spent on ODAs and objectively measured screen time, as recorded in the smartphone settings over the past seven days, were assessed (in hours/week).

The 6-item Facebook Addiction Scale (FAS; Andreassen et al., 2012) and the 6-item Problematic Tinder Usage Scale (PTUS; Orosz et al., 2016) were used to assess problematic use of Facebook and Tinder, respectively, in participants who had utilized these apps in the past 12 months on a 5-point scale.

Mental health

The Beck Depression Inventory-II (BDI-II) was used to assess the severity of depressive symptoms over the past two weeks on a 4-point Likert scale (Beck, Steer, & Brown, 1996; Wintjen & Petermann, 2010). Severity of trait anxiety was assessed with the 20-item State-Trait Anxiety Inventory (STAI-X2) on a 4-point Likert scale (Laux, Glanzmann, Schaffner, & Spielberger, 1981). Attention Deficit Hyperactivity Disorder (ADHD) symptoms were assessed with the 18-item Adult ADHD Self-Report Scale (ADHD-SR) on a 4-point scale (Mörstedt, Corbisiero, & Stieglitz, 2016). Loneliness was assessed using the 3-item Loneliness Scale, measuring the severity of self-perceived loneliness on a 3-point scale (Hughes, Waite, Hawkley, & Cacioppo, 2004). Impulsivity was assessed with the 15-item Barratt Impulsiveness Scale (BIS-15) on a 4-point Likert scale (Meule, Vögele, & Kübler, 2011). Due to a computational error, the first item of the BIS-15 was inadvertently omitted, which should be considered when interpreting the results.

Substance use

Self-reported intensity of alcohol intake was assessed with the 10-item Alcohol Use Disorder Identification Test (AUDIT) on a 5-point Likert scale (Saunders, Aasland, Babor, de la Fuente, & Grant, 1993). General substance use was assessed with a modified version of the Interview for Psychotropic Drug Consumption (IPDC; Quednow, Kühn, Hoenig, Maier, & Wagner, 2004). The 3-month prevalence of cocaine, amphetamine-type stimulants (amphetamine, methamphetamine, MDMA, cathinones), and GHB were assessed. Nicotine use in the past three months was assessed in weekly and daily intake (cigarettes/week/day respectively) and Cannabis use in weekly intake (tetrahydrocannabinol [THC] in grams/week, with one joint accounting for 0.3 g THC), respectively. To explore ODAs role as a potential gateway for substance use, we asked four additional binary questions (y/n): 1) Did you get in contact with illegal substances via ODAs (Either you or your sexual partner used illegal substances on a date); 2) Have Illegal substances been offered to you for sale on ODAs; 3) Did you start taking illegal substances because of ODA use; and 4) Have you increased your use of illegal substances due to ODA use?

Sexual health

Information on STI prevalence and sexual partner count (in the past three months) was obtained from the SwissPrEPared study, corresponding to the most recent visit recorded by the RealRisks study. Lifetime STI prevalence was assessed by a single binary item (y/n). Recent STI prevalence was represented by a positive lab result for chlamydia, gonorrhea, Mycoplasma genitalium, and syphilis within a SwissPrEPared visit (±90 days of the corresponding RealRisks visit). Trait sex motivation was assessed using the 16-item Short Trait Sexual Motivation Questionnaire (s-TSMQ) on a 6-point Likert scale (Stark et al., 2015). Participants were informed that sex/sexual activity referred to both partnered and solitary activity (including masturbation, touching intimate areas, intercourse, and oral sex). Body acceptance and sexual fulfillment were assessed with the respective dimensions of the 35-item Dresden Body Image Questionnaire (DKB-35) on a 5-point Likert scale (Poehlmann, Roth, Brähler, & Joraschky, 2014).

Statistical analysis

Analyses were conducted in R Version 2023.06.1 + 524 (Posit team, 2023). The factorial structure of PODAUS was examined through Structural Equation Modeling (SEM) using lavaan 0.6.17 (Rosseel et al., 2023). In order to investigate PODAUS’ factor structure, an exploratory factor analysis (EFA) was conducted. Confirmatory factor analyses (CFA) were conducted using the Maximum Likelihood estimator with robust standard errors (MLR) to address non-normality. Model fit was evaluated with the Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), Root Mean Square Error of Approximation (RMSEA), and Standardized Root Mean Square Residual (SRMR), adopting the thresholds by Hu and Bentler (1999): CFI and TLI ≥0.90 for acceptable, ≥0.95 for excellent fit; RMSEA and SRMR of 0.08–0.10 as acceptable (Browne & Cudeck, 1992). Reliability and validity were gauged by Cronbach's α (0.80 = excellent, 0.70 = acceptable; Cronbach, 1951) and test-retest reliability with higher values indicating a better fit for temporal stability (Raykov & Marcoulides, 2010). Criterion validity was assessed using Spearman's rank correlations r_s between PODAUS scores and mental health, substance use, and sexual health outcomes using psych 2.4.1 (Revelle, 2024). Associations between PODAUS scores and health outcomes were assessed using multiple linear regression. This analysis employed Ordinary Least Squares (OLS) regression for continuous outcomes, logarithmic regression (LogR) for count outcomes, and Weighted Least Squares (WLS) regression for binary outcomes, and to address heteroscedasticity. Adjustments were made for age and education, and non-normally distributed errors were accepted (Knief & Forstmeier, 2021). To reduce the risk of Type-I error inflation, a significance threshold of p = 0.01 was adopted. Cohen's f² was calculated to interpret effect sizes for continuous outcomes (f² = 0.02 small, f² = 0.15 medium, f² = 0.35 large; Cohen, 1988), Odds ratios (OR) for binary outcomes, and Incidence Rate Ratios (IRR) for count outcomes, respectively (Selya, Rose, Dierker, Hedeker, & Mermelstein, 2012).

Given that our hard-to-reach study population was naturally limited, we did not perform an a priori power analysis. However, we conducted sensitivity power analyses for the main statistical methods (CFA and multiple regression) with the following results: With n = 226 participants, a one-factor solution based on 6 items, a power of 85%, an alpha level of 1%, and an average factor load of 0.65, we can achieve an acceptable CFI of minimum 0.93 in a CFA. When using multiple regression models with up to 10 predictors, a sample size of n = 226, will be sufficient to detect associations with small effect sizes of f² ≥ 0.059, when assuming a power of 85% and an alpha level of 1% (two-tailed).

Ethics

The study procedures were carried out in accordance with the Declaration of Helsinki. The Institutional Review Board of the Ethical Committee in Switzerland approved the study (BASEC-ID 2020-01545) and deemed it a non-interventional, observational study. Participants were informed about the study procedure and provided written informed consent. All study personnel received Good Clinical Practice (GCP) training.

Results

Baseline characteristics

On average, participants spent 9.5 h weekly across 2.7 ODAs, predominantly on Grindr (92%). Among the participants, 54% stated that they got in contact with illegal substances via ODAs, 44% reported that substances were offered for sale on ODAs, 12% reported starting taking illegal substance use because of ODAs, and 23% said their substance use increased due to ODAs. The main motivations for ODA use were to find a sex partner, followed by getting sex fast and chatting.

Tests of dimensionality

CFA results of the German version of the PODAUS demonstrated excellent model fit indices: χ²(9) = 12.92, p = 0.166, CFI = 0.99, TLI = 0.98, and RMSEA = 0.04. Factor loadings ranged from 0.56 to 0.70 (Fig. 2). Information on the EFA can be found in supplementary materials S5 and S6.

Fig. 2.

Factor loadings of the PODAUS

Reliability assessment

PODAUS showed high internal consistency (α = 0.79) and good test-retest reliability (r_s = 0.68) over a 6-month follow-up interval (n = 164), indicating stability over time (see Table 2 and supplementary Fig. S7 for all Spearson's rank correlation and internal consistency results). All other instruments used in this study showed high internal consistency (α > 0.78).

Table 2.

Descriptive statistics, reliability, and Spearman's correlations for all outcome measures with PODAUS score

	M	SD	α	r s	Possible range
Baseline: Problematic Dating App Use (PODAUS)	10.9	3.3	0.79	1.00	6–24
Follow-up: Problematic Dating App Use (PODAUS; N = 164)	10.0	3.1	0.78	0.68**	6–24
Weekly Time spent on Dating Apps (in hours)	9.5	9.5		0.24**
Weekly Smartphone Screentime (in hours; n = 161)	5.1	8.8		0.19*
Age	41.6	11.0		−0.21**
Education	3.8	0.5		0.05
Social Media Use
Problematic Tinder Use (PTUS; n = 75)	8.7	3.2	0.80	0.35**	6–30
Facebook Addiction (FAS; n = 144)	8.3	3.3	0.85	0.22**	6–30
Mental Health
Depression (BDI-II)	7.9	8.8	0.93	0.47**	0–63
Anxiety (STAI-X2)	37.6	11.0	0.94	0.46**	20–80
ADHD (ADHD-SR)	11.7	8.0	0.88	0.37**	0–72
Impulsivity (BIS-15)	27.1	5.5	0.78	0.30**	14–56
Loneliness	1.6	1.6	0.80	0.35**	0–6
Substance Use
Alcohol Use Disorder (AUDIT Score)	5.7	4.2	0.79	0.03	0–40
Nicotine Intake (Cigarettes/day)	3.9	7.7		−0.06
Cannabis Intake (THC in g/week)	0.2	1.0		0.02
3-Month Prevalence Cocaine	0.8	2.2		−0.06
3-Month Prevalence Non-Cocaine Stimulants	0.9	3.6		0.08
3-Month Prevalence GHB	0.6	2.3		0.02
Sexual Health
Lifetime STI Prevalence (Yes, No)	0.7	0.5		−0.22**
3-Month STI Prevalence (n = 202)	0.1	0.3		0.07
Sexual Partner Count	9.4	10.9		0.06
Body Acceptance (DKB-35)	28.0	6.1	0.89	−0.26**	8–40
Sexual Fulfillment (DKB-35)	23.4	4.2	0.85	−0.19**	6–30
Trait Sex Motivation (STSMQ)	2.9	0.8	0.87	0.26**	0–5

Note. *p < 0.05, **p < 0.01, α = Cronbach's alpha, p < 0.01, r_s = Spearman's rank correlation coefficient. Non-Cocaine stimulants include meth-, amphetamines, and cathinones. There were no missing data since all questionnaire fields were set as required, except for assessing external validity in the PTUS (n = 75) and FAS (n = 144), which was only answered by participants if they had used Tinder or Facebook in the past 12 months; and Recent STI Prevalence (n = 202) if there were no STI tests performed in 90± days of the corresponding SwissPrEPared visit.

Validity assessment

PODAUS demonstrated moderate significant positive correlations with the PTUS (r_s = 0.35) and a weak but still significant correlation with the FAS (r_s = 0.22), demonstrating acceptable to good external and content validity (see Table 2 for all Spearman's rank correlation outcomes and supplementary Fig. S7 for a heatmap of Spearman's rank correlation matrix for of PODAUS with all outcomes.). PODAUS correlated significantly positive with weekly time spent on ODAs (r_s = 0.24), demonstrating acceptable construct validity, and weekly overall smartphone screen time (r_s = 0.19), indicating an association with real-life behavior.

Regression outcomes

The cross-sectional multiple regression analysis at baseline accounting for age and education highlighted significant associations of PODAUS with various mental and sexual health outcomes but not with substance use (see Table 3 and full results available in supplementary Table S8). For mental health, higher PODAUS scores were associated with higher symptom scores in depression, anxiety, ADHD, impulsivity, and loneliness, with medium to large effect sizes. For sexual health, significant negative associations were found between PODAUS scores and the lifetime STI prevalence rates (OR: 0.86 [95%-CI 0.78–0.94], body acceptance, and sexual fulfillment with small to medium effect sizes. Conversely, a significant positive association with trait sex motivation was observed with a small effect size. Finally, the PODAUS score did not show a significant relationship with 3-month STI prevalence rates or sexual partner count.

Table 3.

Summary of multiple regression results highlighting only significant effects of PODAUS on outcomes variables (full table can be found in supplementary Table S8)

Outcome	Type	Predictors	B	CI (LL, UL)	R 2	f 2	p
Mental health
Depression	WLS	PODAUS	1.00	0.70, 1.30	0.24	0.32	<0.001
		Age	−0.07	−0.14, 0.01			0.08
		Education	−2.96	−5.34, −0.58			0.015
Anxiety	WLS	PODAUS	1.39	1.01, 1.78	0.30	0.43	<0.001
		Age	−0.22	−0.33, −0.11			<0.001
		Education	−3.72	−6.51, −0.94			0.009
ADHD	WLS	PODAUS	0.82	0.52, 1.12	0.17	0.21	<0.001
		Age	−0.11	−0.20, −0.03			0.010
		Education	0.10	−1.80, 2.01			0.914
Impulsivity	WLS	PODAUS	0.51	0.30, 0.72	0.15	0.18	<0.001
		Age	−0.08	−0.14, −0.02			0.012
		Education	−1.20	−2.61, 0.22			0.097
Loneliness	OLS	PODAUS	0.17	0.11, 0.23	0.20	0.25	<0.001
		Age	−0.00	−0.02, 0.02			0.831
		Education	−0.91	−1.30, −0.53			<0.001
Sexual health
Lifetime Prevalence STI	LogR	PODAUS	−0.15	0.78, 0.94	0.06	0.86a	0.001
		Age	0.01	0.98, 1.03		1.01	0.652
		Education	0.06	0.58 1.88		1.07	0.826
Sex Motivation	OLS	PODAUS	0.07	0.04, 0.11	0.09	0.09	<0.001
		Age	0.01	−0.00, 0.02			0.169
		Education	0.07	−0.15, 0.28			0.546
Body Acceptance	OLS	PODAUS	−0.47	−0.70, −0.23	0.14	0.16	<0.001
		Age	0.10	0.03, 0.17			0.006
		Education	1.76	0.22, 3.30			0.025
Sexual Fulfillment	WLS	PODAUS	−0.21	−0.38, −0.05	0.10	0.11	0.010
		Age	0.08	0.03, 0.12			0.001
		Education	0.60	−0.57, 1.77			0.312

Note. LL = Lower limits, UL = Upper limits, WLS = Weighted least squares, OLS = Ordinary least squares, LogR = Logarithmic regression, ^a = Odd's ratio, significant outcomes (p < 0.01) are highlighted in bold.

Discussion and conclusion

This study aimed to investigate the association between problematic ODA use and mental and sexual health outcomes within an MSM-cohort counseling for HIV PrEP by initially validating a German version of the PODAUS, a brief innovative instrument developed to quantify problematic ODA use.

Consistent with the original Italian version of the PODAUS (Gori et al., 2024), the German version confirmed a unifactorial 6-item structure through CFA analysis, showing excellent psychometric properties. It demonstrated high internal consistency and good test-retest reliability at 6 months. Comparisons between the German and Italian versions indicated both methodological consistencies and differences. Both versions identified similar factor structures and loadings. Despite the scale modifications (excluding a neutral option to minimize response bias and item differentiated regarding mood modification and salience), the German PODAUS remains reliable and valid, suggesting further exploration into these differences. Notably, linguistic and demographic variations existed between the studies; Gori et al. (2024) primarily included young adult heterosexual females, whereas our study focused on mid-aged MSM in Switzerland, potentially influencing ODA usage patterns. Cultural distinctions likely influenced ODA usage patterns, with Swiss MSM sample showing longer daily usage durations compared to the Italian convenience sample. The consistent findings in the Italian and MSM samples indicate that the PODAUS is a valid and reliable scale measuring problematic ODA use across two diverse groups. However, future investigations should analyze possible measurement invariance across different cultures and diverse demographics.

Convergent validity was established through Spearman's rank correlations and multiple regression analyses, showing significant links with mental and sexual health outcomes. Significant positive correlations of PODAUS with other behavioral addiction scales like the PTUS and FAS demonstrated acceptable external validity, indicating problematic use across various platforms and possible underlying psychological issues. This was further supported by significant positive, even though weak, correlations with weekly ODA use time and total smartphone screen time, suggesting that problematic ODA use is associated with more behaviors that reflect a loss of control in user interactions. These results align with the hypothesis that principles of social networking site addiction are relevant to ODA research (Bonilla-Zorita, Griffiths, & Kuss, 2021).

Multiple regression analyses showed that problematic ODA use is significantly negatively associated with mental and sexual health. PODAUS was a significant predictor of symptoms of depression, anxiety, loneliness, impulsivity, and ADHD, supporting previous findings on the adverse effects of ODA use on MSM's mental health (review, by Obarska, Szymczak, Lewczuk, & Gola, 2020). The link between depressive and anxious symptoms and PODAUS could arise from neglecting positive activities, increased self-comparison, and the attractiveness of ODAs to those experiencing anhedonia (Lin et al., 2016; Shensa et al., 2017; Steers, Wickham, & Acitelli, 2014). Negative ODA experiences like exclusion and lack of social connection, along with a “fear of missing out,” may also heighten the psychiatric burden, particularly in MSM (Bonilla-Zorita et al., 2021; Gupta & Sharma, 2021; Kuss & Griffiths, 2017).

The lack of significant associations between problematic ODA use and substance use indicates that its psychological effects are more closely tied to mental and sexual health than to substance behaviors. This is consistent with literature that usually associates substance use with casual sexual encounters rather than with ODA use, specifically (Choi, Wong, & Fong, 2017). This disparity might also reflect a different vulnerability between substance use and behavioral addictions, given their distinct etiopathogeneses (Starcevic & Aboujaoude, 2017). However, a subgroup of participants reported to have initiated or increased their substance use due to their ODA use, indicating that ODAs may function as a gateway for substance use. These findings, although only cross-sectional, emphasize the need for identifying those most vulnerable to prevent long-term adverse outcomes, such as the development of substance use disorders (Boonchutima & Kongchan, 2017). Longitudinal studies are needed to further investigate the association between technology-, behavior-, and substance-related addictive behaviors.

PODAUS significantly predicted diminished body satisfaction, reduced sexual fulfillment, and elevated trait sex motivation. Yet, surprisingly, it was linked to lower lifetime STI prevalence but not with recent STI prevalence or partner counts. This finding contrasts with studies that show increased sexual risks among ODA users (Choi et al., 2016; Flesia et al., 2021b; Lehmiller & Ioerger, 2014; Liberacka-Dwojak et al., 2023; Sawyer, Smith, & Benotsch, 2018) but aligns with research showing no significant differences in sexual behaviors between ODA users and non-users (Andersson, Preuss, Boman, & Nylander, 2019; Wu, 2019). Notably, problematic ODA use was not significantly associated with partner count. Despite high levels of trait sexual motivation, some users may remain active on ODAs without ever meeting a potential partner. This is reflected by the significant association of high PODAUS scores, alongside low body satisfaction and low sexual fulfillment. This may help explain the absence of a link with recent STI prevalence. Mental health challenges, such as elevated depressive and anxious symptoms observed in our cohort, could further impact sexual health outcomes. The absence of significant effects between higher PODAUS scores and STI prevalence rates may be attributed to differences in sexual health practices in our participants, who are part of the test-and-treat SwissPrEPared program, which includes regular tests and counseling, as observed in other studies. This may contrast with MSM receiving only standard care, as literature suggests that retention in care has protective effects on health outcomes (Collins et al., 2021; Winter et al., 2022). More longitudinal research is needed to understand the interplay of problematic use and sexual health, including the analysis of protective factors.

One strength of the PODAUS is its multi-app approach, capturing data on which ODA was utilized and the primary motivation for its use, reflecting the complex use of multiple apps simultaneously, responding to current ODA trends, and catering to each person's needs, which should be kept in mind for future researchers when assessing problematic ODA use. This broadens the traditional scope of dating apps to include non-traditional platforms like Instagram, indicating that the focus should be on user interactions with specific ODAs rather than the apps themselves. Moreover, our study evaluated the test-retest reliability of the PODAUS through two time points, and used a comprehensive set of validated questionnaires for assessing mental and sexual health.

This study has some limitations that warrant consideration. The sample, primarily well-educated and financially stable MSM born in Switzerland, exhibits selection bias, probably skewing our results and affecting the generalizability, as previously noted (Winter et al., 2022). Future research should strive to include participants from a broader range of educational, socio-economic, and cultural backgrounds, and include longer follow-up periods, to enhance the generalizability of the findings.

Self-reported data, particularly on sensitive topics like substance use and sexual behavior, may be prone to reporting biases underreporting and inaccuracies (Hampel et al., 2020). This is supported by two studies demonstrating that underreporting of recreational substance use rates ranges between 30 and 60% (Janousch et al., 2024; Steinhoff et al., 2023). Future research should use objective measures such as screentime tracking and onsite tests for substance use and STIs to validate responses and mitigate social desirability and recall biases (Hussy, Schreier, & Echterhoff, 2013). Additionally, external factors like the COVID-19 pandemic (2020–2022) and the MPOX outbreak (summer 2023) could have influenced participant behaviors and health outcomes, introducing challenges related to mental health, social isolation, and economic difficulties (Studer et al., 2021; Winter et al., 2022).

It's crucial to understand how different ODAs and their features affect mental health symptoms in users and to clarify when problematic use escalates to addiction. Future research should differentiate between passionate and problematic use to prevent the overpathologizing of everyday behaviors (Cataldo, Billieux, Esposito, & Corazza, 2022; Flayelle, Schimmenti, Starcevic, & Billieux, 2022; Fournier et al., 2023). Our qualitative interviews as well as the PODAUS results indicate that problems with ODA use seem to be common in our study population. Specifically, the association between problematic ODA use and poorer mental health suggests that it may serve as a warning sign for clinicians to further assess mental health concerns, and similarly, mental health issues could indicate a need to evaluate ODA use. This reciprocal relationship underlines the necessity for targeted prevention and intervention strategies. Longitudinal research is essential for understanding the dynamics of this relationship and evaluating intervention efficacy. Currently, there is a gap in care for individuals with problematic ODA use, highlighting the need for integrated care models that address problematic ODA use, mental health, substance use, and sexual health to better meet the complex needs of vulnerable populations, such as MSM. In a next step, collaborations among mental health professionals, ODA developers, and community organizations are essential for launching awareness campaigns and developing support tools. These initiatives should focus on educating users about responsible ODA use, digital literacy, recognizing signs of misuse, and promoting periodic breaks from app usage. Considering the more significant negative impact on mental health among sexual minority communities, tailored mental health support for these groups is vital. A deeper understanding of ODA use dynamics will help craft public health interventions that minimize risks while enhancing the social benefits of ODAs.

In conclusion, the 6-item PODAUS emerges as a valid and reliable self-report scale for assessing problematic ODA use. Its development provides a valuable tool for researchers and clinicians to evaluate and monitor problematic ODA usage, which was associated with adverse mental, sexual, and physical health outcomes in the present study.

Supplementary material

jba-14-178-s001.pdf ^{(405.8KB, pdf)}