key: cord-0896994-vn1o3dnv
authors: Gleason, Neil; Banik, Swagata; Braverman, Jesse; Coleman, Eli
title: The Impact of the COVID-19 Pandemic on Sexual Behaviors: Findings from a National Survey in the United States
date: 2021-08-23
journal: J Sex Med
DOI: 10.1016/j.jsxm.2021.08.008
sha: 29c1a7990d296ed35f27d7e0cb671546466e4576
doc_id: 896994
cord_uid: vn1o3dnv

Background: Studies from the first months of the COVID-19 pandemic and the resulting lockdown and social distancing measures have shown that there have been decreases in sexual frequency and relationship satisfaction. Aim. To evaluate the ongoing impact of the COVID-19 pandemic on sexual behavior, relationship satisfaction, and intimate partner violence in the U.S. using a large national convenience sample. Methods. 1,051 participants across the U.S. were recruited in October 2020 to complete a cross-sectional online survey. Outcomes: Participants were asked to retrospectively report their sexual behavior frequency, relationship satisfaction, and intimate partner violence during the pandemic and prior to the pandemic Results. There was a small but significant decrease in some retrospectively-reported partnered sexual activities, and men reported a small increase in masturbation and pornography use. There was no evidence for a change in relationship satisfaction or intimate partner violence, but both men and women reported a small decrease in sexual pleasure, and women reported a small decrease in sexual desire. The sexual behaviors with greatest reduction were casual sex, hookups, and number of partners, and the most diminished as aspect of sexual functioning was sexual enjoyment. Depression symptoms, relationship status, and perceived importance of social distancing emerged as predictors of these reductions. Less than half of individuals who engaged with casual sex partners before the start of the pandemic ceased this behavior completely after the start of the pandemic. Individuals waited on average 6-7 weeks before reengaging in casual sex. Clinical translation: These results inform public health response to the effects of the pandemic and add to our understanding of how the pandemic has continued to impact sexual behavior. Strengths and Limitations: This is the first known study to evaluate sexual behavior several months into the COVID-19 pandemic using a large national sample. However, the results of this study are limited by its convenience sampling method and cross-sectional design. Conclusions: These results indicate that the changes in sexual behavior observed in the early months of the pandemic have continued, with small but significant decreases in many partnered sexual behaviors and a small increase in men's solitary sexual behaviors.

To evaluate the ongoing impact of the COVID-19 pandemic on sexual behavior, relationship 21 satisfaction, and intimate partner violence in the U.S. using a large national convenience sample. 22

Methods. 1,051 participants across the U.S. were recruited in October 2020 to complete a cross-23 sectional online survey. 24

Outcomes: Participants were asked to retrospectively report their sexual behavior frequency, 25 relationship satisfaction, and intimate partner violence during the pandemic and prior to the 26 pandemic 27

Results. There was a small but significant decrease in some retrospectively-reported partnered 28 sexual activities, and men reported a small increase in masturbation and pornography use. There 29 was no evidence for a change in relationship satisfaction or intimate partner violence, but both men 30 and women reported a small decrease in sexual pleasure, and women reported a small decrease in 31 sexual desire. The sexual behaviors with greatest reduction were casual sex, hookups, and number 32 of partners, and the most diminished as aspect of sexual functioning was sexual enjoyment. 33

Depression symptoms, relationship status, and perceived importance of social distancing emerged 34 as predictors of these reductions. Less than half of individuals who engaged with casual sex 35 partners before the start of the pandemic ceased this behavior completely after the start of the 36

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was recognized in 59 January 2020 as the agent responsible for the coronavirus disease (COVID-19), whose worldwide 60 outbreak led to the declaration of a global pandemic by the World Health Organization (WHO) on 61

March 11, 2020 1 . In response to the COVID-19 pandemic, governments across the globe have 62 who reported any physical or sexual violence were asked if they had experienced an increase in 157 sexual, physical, or emotional violence since the start of the COVID-19 pandemic. Finally, 158

participants were asked if they had physically threatened a partner or forced a partner to have sex 159 during the COVID-19 pandemic or in the year before the pandemic. 160

Participants were asked to report relevant demographic, substance use, and COVID-19 162 exposure information (see Table 1 ). In addition, participants completed an adapted version of the 163 Patient Health Questionnaire-2 (PHQ-2) 17 , which asked about depression symptoms in the past two 164 weeks and in the year before the pandemic. 165

The survey included 5 attention check questions. Participants were asked for their age 167 twice during the survey to check for discrepancies and were asked 3 multiple choice questions with 168

only one correct answer: ii) "If you are reading this question, please select 'sometimes.'"; ii) "what 169 color is the sky?"; and iii) "how often is W the first letter of the alphabet?" (with the correct answer 170 being "never"). 171

Descriptive statistics and effect sizes were calculated for sexual frequencies and measures 173 of sexual and relationship functioning reported before and after the start of the COVID-19 pandemic 174 Differences in these scores were also assessed with a series of within-subject t-tests. A Bonferroni 175 correction was applied to account for familywise error in multiple statistical tests, resulting in a 176 critical value of p=.0038. To assess demographic correlates of substantial changes in sexual 177 frequency of functioning (i.e. small effect sizes or larger), a series of multiple linear regressions 178 were conducted. Demographic correlates included in the model were: gender (male vs. female), 179 sexual orientation (heterosexual vs. non-heterosexual), age, relationship status (partnered vs. non-180 partnered), political affiliation, religion (religiously affiliated vs. agnostic/atheist/none), whether 181 they have children at home (yes vs. no), financial difficulties due to COVID-19 (yes vs. no), 182 perceived importance of following social distancing guidelines, and depression symptoms (PHQ-2 183 score). 184

Differences in pre-and post-pandemic sexual frequency and satisfaction scores, including 186 effect sizes of retrospectively-reported changes, are displayed in Table 2 . Because only 19 187 participants reported having sex with a romantic partner their current partner was not aware of 188 (i.e. an "affair"), this question was excluded from analysis. A series of within-subject t-tests 189 indicated several significant but very small (d<.2) differences: a significant increase in frequency of 190 masturbation and pornography use, and a significant decrease in frequency of sex with current 191 partner and in ratings of sexual desire. Small significant decreases (d>.2) were noted for number of 192 sex partners, frequency of hookups, and ratings of sexual enjoyment/pleasure, and a medium 193 significant decrease (d>.5) was noted for frequency of sex with casual partners (see Table 2 ). These 194 analyses were repeated for men and women separately (see Table 3 ) and some gender differences 195 were noted. Men reported very small (d<.2) but significant increases in pornography use and 196 masturbation while women did not. In addition, women reported a very small (d<.2) but significant 197 decrease in sexual desire, while men did not. 198

Three multiple linear regressions were run to assess demographic correlates of three sexual 199 behavior changes that small to medium Cohen's D effect sizes in the series of within-subject t-tests 200 (number of sexual partners, sexual enjoyment/pleasure, and sex with casual partners). Though 201 changes in hookups demonstrated a small effect size (d=.46), the sample size (N=68) was 202 insufficient to include this variable in a multivariate analysis. Results are displayed in Table 4 . 203

Higher depression scores significantly predicted decreased sexual enjoyment/pleasure and casual 204 partner frequency. Greater perceived importance of social distancing was associated with 205 decreased casual partner frequency and being single was associated with a greater decrease in 206 number of sexual partners. 207

To better understand how changes in sex with current partner differed based on 208 relationship status, differences in means were compared for participants who were married, in a 209 live-in relationship, and in a relationship but living separately. The mean difference for those in a 210 relationship and living separately (N=130) was ∆M=-.445, which was a small effect size (d=.242). 211

The mean differences were much smaller for participants who were in a live-in relationship 212 (N=148; ∆M=-.049; d=.054) or married (N=421; ∆M=-.082; d=.095). This indicates that the small but 213 statistically significant reduction in sex with current partner was driven by those living separately 214 from their partners. 215

Of participants who engaged in at least one hookup in the year before the start of the 216 pandemic (N=60), 24 (40.0%) did not engage in any hookups after the start of the pandemic. 217

Participants who engaged in hookups during the pandemic (N=42) reported waiting an average of 218 6.67 (SD=5.85) weeks after the start of the pandemic before having sex with a new partner. Thirty 219 individuals that engaged in hookups (44.1%) reported using a condom every time, and 55 (80.9%) 220 reported that they met at least some of their partners on a dating/hookup website or app. Similarly, 221 of participants who engaged in sex with a casual partner at least once in the year before the start of 222 the pandemic (N=179), 65 (36.31%) reported no sex with a casual partner during the pandemic. 223

Participants who reported engaging in sex with a casual partner during the pandemic (N=125) 224

reported waiting an average of 6.26 (SD=6.89) weeks before engaging in sex with a casual partner. 225

Seven of these individuals (5.6%) reported that they currently live in the same household as their 226 casual partner. Ninety-two individuals that reported sex with a casual partner (47.7%) reported 227 using a condom every time. One hundred nineteen participants reported that they were currently in 228 a relationship but living separately, and of these, 15 (12.61%) reported that they had not had sex 229 with their partner during the pandemic. 230

The mean number of sex partners reported during the pandemic was M=1.0 (SD=1.46), 231 compared to a mean of M=1.27 (SD=1.72) in the 6 months before the start of the pandemic. 232

Participants that reported having multiple sex partners either before or during the pandemic 233 (N=209) reported an average of M=2.16 (SD=2.87) partners during the pandemic and an average of 234 M=3.27 (SD=3.02) partners in the 6 months before the start of the pandemic. The effect size for the 235 difference between these means was small (d=.441). 236

Sexual and physical violence was reported by a small number of participants (see Table 5 ), 237 and therefore statistical analyses were not conducted to determine differences in reported violence 238 before and during the pandemic. Among those who reported being the victim of any sexual or 239 physical violence before or during the pandemic (N=61), 17 (27.9%) reported that they had 240 experienced more physical, sexual, and emotional violence during the COVID-19 pandemic, while 241 44 (72.1%) did not report an increase. 242

This is the first known study to examine sexual behavior six months into the COVID-19 244 pandemic in the United States using a large sample. The observed small decreases in partnered 245 sexual frequency and sexual functioning, as well as the small increase in solitary sexual frequency, 246 are largely in line with studies conducted during the first months of the pandemic. It was also noted 247 that less than half of participants who reported engaging in casual sex before the pandemic 248 reported stopping this behavior completely during the pandemic. Those who continued to engage 249 with outside sexual partners waited on average 6-7 weeks until after the start of the pandemic 250 before they began engaging the behavior again. Thus, while casual sex may have declined sharply at 251 the start of the pandemic, as was observed in previous studies 9,10,12,13 , frequency of these behaviors 252 may have rebounded after a few months. 253

Sexual health, frequency, and satisfaction impacts relationship satisfaction and overall 254 mental and physical well-being 18-21 , and thus the small and medium retrospectively-assessed 255 changes in sexual behavior may indicate an emerging public health concern. However, from the 256 research that has been conducted during the COVID-19 pandemic, it remains unclear whether 257 reduced sexual satisfaction and pleasure are impacting mental health or whether poorer mental 258 health is impacting sexual satisfaction. This study, for instance, found that depression symptoms 259 are related to decreases in sexual enjoyment/pleasure and casual sex, though the direction and 260 causality of this relationship is unclear. Mental and sexual health appear to have declined 261 simultaneously during the COVID-19 pandemic 22 , and thus could be impacting each other or having 262 a combined negative impact. 263

Overall, it is encouraging news that reduction in sexual desire and frequency among 264 married and cohabiting partners was modest, given the strain the pandemic has caused on partners 265 living and oftentimes working in the same space for long periods of time. Understandably, the 266 negative impact was worse among partners living apart. While the methodological limitations of 267 this study prevent a more detailed analysis of the data on sexual and domestic violence, it at least 268 provides an encouraging, albeit tentative, indicator that it has not overwhelmingly increased , 269 which was been feared and reported early on in the pandemic 6 . Since the social distancing 270 requirements have been and may continue to be in effect for some time, the long-term impact on 271 relationships is still unclear. 272

The reduction in sexual activity among non-partnered and young adults is also a matter of 273 concern. Over the past several decades, sexual frequency among young adults has declined in the 274

United States 23-27 and some other developed countries 28 . Whether this decline is associated with 275 trends of poorer mental health among youth is not clear as there are other factors that have been 276 implicated (e.g, increased use of technology, reduced sleep) 26 . It is not yet known whether the 277 COVID-19 pandemic will exacerbate this trend, or whether its effect will be temporary or 278 permanent. 279

Several limitations must be considered. First, the cross-sectional nature of the study limits 281 the causal conclusions that can be drawn. Because individuals were asked to retrospectively report 282 their sexual behavior from months prior to the survey, their answers may have been influenced by 283 recall bias. Ideally, data would have been collected in the months before the pandemic began as a 284

comparison. In addition, most variables were assessed with single-item measures that were 285 developed for the purpose of the study, which limits the ability to compare these data to other 286 studies conducted in different time periods and settings. 287

Second, while the sample represented a wide range of demographics, it was a convenience 288 sample collected through Mturk panel and was not intended to be a fully representative sample. It 289 was noted that compared to 2019 US census estimates 29 , this sample had a larger proportion of 290 males and white participants, and had fewer Hispanic and black participants. In addition, 291 participants in this sample were more highly educated, less religious, and more politically liberal 292 than the general US population: 49.8% of our sample had obtained a bachelor's degree compared to 293 35.0% nationally 30 , 51.7% identified as religiously unaffiliated (e.g. agnostic, atheist, or none) 294 compared to 22.8% nationally 31 , and 56.0% identified as politically liberal compared to 26% 295 nationally 32 . There were also a large portion of surveys (17.0%) that were excluded from analysis 296 because of failed attention check questions, which is indicative of a well-documented problem with 297 data quality on the Mturk platform 33 . However, other studies using Mturk to recruit participants for 298 public health research have found the samples they recruited to be adequately representative of the 299 wider U.S. population 34 , and the attention check questions used for this survey helped to ensure the 300 final data used in analysis was valid. 301

Future research should further explore whether changes in sexual frequency were more 302 pronounced or diminished among certain demographics such African American and Latino 303 communities, which have experienced sharp disparities in rates of COVID-19 cases and deaths . It 304 would also be useful to further examine personality, sociocultural, and structural factors that may 305 play a role in these changes. For instance, individuals with personality characteristics associated 306 with high sexual frequency such as sexual compulsivity, sexual sensation seeking, or sexual 307 disinhibition, may have experienced less reduction in sexual frequency compared to other 308

individuals. In addition, sociocultural and structural factors that have come into play during the 309 pandemic, such as financial hardship, discrimination, and geography, also likely had an impact on 310 changes in sexual behavior. Collecting data on these factors would assist with targeted outreach to 311 individuals who are most at risk of continuing unsafe behavior during the pandemic. Finally, the 312 majority of research published to date has examined sexual behavior during the first few months of 313 the pandemic, and our study found that six months into the pandemic, sexual behavior looked quite 314 different. Continuing to track sexual behavior throughout the course of the pandemic will allow 315 research to identify high-risk groups during pandemic and inform targeted public health 316

interventions. 317

Future research should also address pandemic preparedness and response as it relates to 318 sexual and reproductive health services. Because the COVID-19 pandemic may have long term 319 implication on birth rate, post-pandemic sexual behavior, and psychological health, research should 320 examine ways to respond to and mitigate these negative outcomes. It is also important to consider 321 strategies for training the next generation of sexual health therapists and healthcare providers, as 322 they will be on the front lines of addressing these issues across different demographic groups and 323 localities. Note. *significant at p<.0038 (Bonferroni correction applied). Those identifying as "other" gender 486

were excluded due to low sample size. a Mean differences were calculated by subtracting current 487 frequency/ratings from frequency/ratings before the COVID-19 pandemic. Negative mean 488 differences indicate a decrease in frequency/ratings, positive mean differences indicate an increase 489 in frequency/ratings. b 13 within-subjects t-test were conducted. c Affair data were not included in 490 the analyses due to low frequency. d Only participants that reported being in the same romantic 491 relationship before and after the start of the COVID-19 pandemic were included in analysis. 492 493 

Timeline: WHO's COVID-19 response

Making love in the time of corona -considering relationships in 340 lockdown

The Impact of COVID-19 on Sexual 342 Health: A Preliminary Framework Based on a Qualitative Study With Clinical Sexologists

Sexual and reproductive health (SRH): a key issue in the 345 emergency response to the coronavirus disease (COVID-19) outbreak. Reprod Health

Ten Days to Self-Esteem: The Leader's Manual. Harper Collins; 1993. 383 16. Domestic Violence Risk Assessment

Detecting and monitoring depression with a two-item 386 questionnaire (PHQ-2)

Promoting sexual health and responsible sexual behavior: an introduction

Sexuality and Health among Older Adults in the United States

Sexual Health for the Millennium: A Declaration and Technical Document

Psychiatry and Sexual Medicine: A Comprehensive Guide for Clinical 397 Practitioners

Mental Distress in the United States at the Beginning of 399 the COVID-19 Pandemic

Sexual Activity and Contraceptive Use Among Teenagers in the United 402 States

Youth Risk Behavior Surveillance -United States

Explaining the Decline in Young Adult Sexual Activity in the United States

Age, period, and cohort trends in mood 408 disorder indicators and suicide-related outcomes in a nationally representative dataset

Trends in Frequency of Sexual Activity and 411 Number of Sexual Partners Among Adults Aged 18 to 44 Years in the US