Love during lockdown: findings from an online survey examining the impact of COVID-19 on the sexual practices of people living in Australia

Introduction: Australia recorded its first case of COVID-19 in late January 2020. On 22 March 2020, amid increasing daily case numbers, the Australian Government implemented lockdown restrictions to help flatten the curve. Our study aimed to understand the impact of lockdown restrictions on sexual and reproductive health. Here we focus on sexual practices. Methods: An online survey was open from the 23 April 2020 to the 11 May 2020. Participants were recruited online via social media and other networks and were asked to report on their sexual practices in 2019 and during lockdown. Logistic regression was used to calculate the difference (including 95% confidence intervals) in the proportion of sex practices between time periods. Results: Of the 1187 who commenced the survey, 965 (81.3%) completed it. Overall 70% were female and 66.3% were aged 18 to 29 years. Most (53.5%) reported less sex during lockdown than in 2019. Compared with 2019, participants were more likely to report sex with a spouse (35.3% vs 41.7%; difference=6.4%; 95%CI: 3.6, 9.2) and less likely to report sex with a girl/boyfriend (45.1% vs 41.8%; diff=-3.3%; 95%CI: -7.0, -0.4) or with casual hook-up (31.4% vs 7.8%; 95%CI:-26.9, -19.8). Solo sex activities increased, 14.6% (123/840) reported using sex toys more often and 26.0% (218/838) reported masturbating more often. Dating app use decreased during lockdown compared with 2019 (42.1% vs 27.3%; difference= -14.8%; 95%CI: -17.6, -11.9). Using dating apps for chatting/texting (89.8% vs 94.5%; diff=4.7%; 95%CI:1.0, 8.5) and for setting up virtual dates (2.6% vs 17.2%; diff=14.6%; 95%CI:10.1, 19.2) increased during lockdown. Conclusion: Although significant declines in sexual activity during lockdown were reported, people did not completely stop engaging in sexual activities during the pandemic, highlighting the importance of ensuring availability of normal sexual and reproductive health services during global emergencies.


INTRODUCTION
In late January 2020, Australia recorded its first case of COVID-19 [1]. In response to rapidly increasingly daily case numbers [1], the Australian Government began to introduce several measures in an attempt to 'flatten the curve'. On the 22 nd March, Stage 1 restrictions were announced, including the temporary closure of non-essential businesses and services, limiting the size of non-essential gatherings, promoting social distancing and advising against nonessential travel [2]. As daily case numbers continued to top 300 in the following week [1], strict lockdown measures were implemented from the 29 th March, with people asked to remain in their homes and only leave for four activities: shopping for essential goods and services, to exercise, to seek medical care or to attend work or education where these activities could not take place at home [3]. Gatherings were limited to two persons only [4]. Children were homeschooled by their caregivers, and large swathes of the workforce began working from home indefinitely. In addition to the closure of the international border, most state and territory borders were also closed, effectively preventing interstate travel (and at the time of writing, have not yet reopened). These strict lockdown measures continued until the 8 th May 2020, when the Australian Government announced a plan for the easing of restrictions and a 'COVID-safe Australia' [5] as case numbers across the country consistently declined to below 30 1 a day [1].
During this isolation period, colloquially referred to by the Australian public as 'iso', most people significantly reduced their activities outside their homes [6], and a survey conducted by the Australian Bureau of Statistics found 94% of participants were keeping their distance from those outside their household [7]. Intimate relationships were not exempt from the impact of the lockdown. As restrictions were enforced across the country, confusion abounded among those with non-cohabitating intimate partners, as they wondered whether visits were allowed under the new restrictions. While some states banned people, including non-cohabitating intimate partners, to meet unless they were exercising together or providing care [8], others changed course during lockdown and ultimately made an exception to allow non-cohabitating partners to meet [9,10]. For those isolating with their partners, some prophesised that Australia is much more likely to experience a declining birth rate than a baby-boom nine months postlockdown [11] as people grapple with the social, economic and health impacts of the pandemic. For those without regular partners, sexual health organisations and state health departments advised against casual sex during lockdown but provided advice on how to still enjoy sexual pleasure in the absence of physically present partner/s [12][13][14]. Dating apps like Tinder and Hinge included in-app safety messages about how to connect with new partners while maintaining social distancing, suggesting using video platforms to set up virtual dates [15].
Evidently, the COVID-19 pandemic and the lockdown it prompted are likely to have an impact on the sexual practices and sexual health of people living in Australia. We implemented a serial cross-sectional survey that aimed to investigate the impact of COVID-19 on sexual and reproductive health of people living in Australia. In this paper we report on the results from the first survey and explore the impact of lockdown on people's sexual practices.

Sexual and Reproductive Health during COVID-19 Online Survey
The first survey was administered online between the 23 rd April to 11 th May. All Australian states and territories were under lockdown to varying degrees during this period, with restrictions easing from the 8 th May. The survey comprised questions pertaining to the impact of COVID-19 on sexual practices and sexual and reproductive health. As a repeated crosssectional survey, participants were asked to provide their month and year of birth and their 'porn star name' (name of their first pet and the name of the first street they lived on) to enable participant tracking over time [16]. Participants were also asked to provide an email address if they wished to be contacted for future surveys. Email addresses were removed from the dataset and not used to link responses. Repeat waves of the survey will be administered every six to eight weeks across the remainder of 2020, and a cohort analysis will be undertaken on those participants who respond more than once. In this paper, we report on data pertaining to the baseline survey only (subsequent waves of data collection are ongoing). This study was approved by the University of Melbourne Human Research Ethics Committee (ID: 2056693).

Participants
Participants were recruited via various means, including the research team's existing networks (for example, emailing the recruitment flyer to colleagues and sexual and reproductive health newsletter lists for distribution, posting on student noticeboards and personal social media accounts and posting the recruitment flyer on our research group's Twitter account) and via paid Facebook ads. Participants were asked to pass the link to the survey on to their own networks. People were eligible to participate if they were aged 18 years or older and living in Australia at the time of the survey. Participants clicked on a link that took them to the survey page where they were provided with a plain language statement. If they were interested in participating, they were asked to confirm that: they were aged 18 years or over, they understood what the survey was about, and they consented to participate.

Data collection
Survey questions included trends and changes in sexual practices, intimate relationships, access to essential goods and services, trends and changes in contraceptive use and pregnancy intentions. Sex was defined as physical contact with other people for sexual pleasure including oral sex and mutual masturbation. For several questions, participants were asked to report on two time periods: their experiences and practices during all of 2019 and during 'lockdown' (the period after March 22 nd when restrictions commenced). Participants were also asked to report on the frequency of sexual activity by commenting on whether their sexual activity was less, the same or more during lockdown than in 2019. They were also asked to report whether specific sexual practices (such as masturbating alone or oral sex) were being performed the same amount, less often, more or stopped completely because of COVID-19. We also collected sociodemographic data. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 12, 2020. . https://doi.org/10.1101/2020.08.10.20171348 doi: medRxiv preprint

Data analysis
A sample size of 800 would allow us to detect a difference in paired proportions of 6% (55% vs 56%) assuming a correlation of 0.25, power of 80% and alpha of 0.05. Descriptive statistics were used to describe the sociodemographic characteristics of participants. Logistic regression was used to determine the difference in proportion between the two time periods (lockdown minus 2019) for categorical variables adjusting for clustering at the participant level. The difference in the proportion and its corresponding 95% confidence intervals are reported. We fitted interaction terms between the time period (lockdown versus 2019) and the variables of gender, age, sexuality and relationship status in each logistic regression model to examine whether the difference in reported activities between time periods varied across different categories of these variables (for example, did the difference in app use between the two time periods vary between those aged 18 to 29 years and those aged 30+ years). We also used Chi square tests and tests for equality of proportions to investigate associations between categorical variables where indicated and t or Mann Whitney tests to compare continuous variables between two groups. As not everyone completed all questions, missing data are excluded from all analyses, but denominators are provided to put these missing data into context. To assess response bias, we compared the gender, age and sexuality of those who completed and did not complete the survey. All statistical analyses were performed using Stata SE 16.0 for Windows.

Demographics
A total of 1187 people consented to participate and commenced the survey; 965 (81.3%) completed it and were included in the analysis. Those who completed the survey were older on average than those who did not complete (28.6 vs 26.2 years, p<0.01) but there was no difference in gender (p=0.44) or sexuality (p=0.11). Overall, 70.0% were female, 66.3% were aged under 30 (median age=24; IQR=20-33; range 18-77), 82.7% reported Australia as their country of birth and 65.7% indicated they were heterosexual. In 2019, 80.2% reported being employed, but this fell to only 63.4% during lockdown. Overall, 9.1% reported testing for COVID-19, but none had tested positive. Sociodemographic data are reported in Table 1. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 12, 2020. N=Number who answered the question, denominator is not always 965 because of missing data; * Gender diverse includes transgender and non-binary; † 'In a relationship' includes those who selected married/de facto/boyfriend/girlfriend/LAT (living apart together) or long distance relationship; 'cohabitating' defined as living with partner/s; 'Single' includes those who selected single/divorced/widowed (and did not select another relationship); 'Other' includes people who selected polyamorous or multiple partners (and did not select another relationship option, such as single), and those who indicated in comments that they were casually dating, but not exclusive; ‡ MSM includes bisexual and trans men; WSW includes bisexual and trans women; Other includes asexual and non-binary respondents, and those who selected 'something else' ; § Participants can be in multiple categories; || Includes retired, parent/carer, disability support pension, unemployed but not looking for work; ¶ Excluding students

Sexual activity
Participants reported a median of one sex partner in 2019, (IQR 1-3; range: 0-1000) and a median of one sex partner (IQR 0-1; range: 0-10) during lockdown. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 12, 2020. . https://doi.org/10.1101/2020.08.10.20171348 doi: medRxiv preprint sexuality (p<0.01) and by relationship status (p<0.01). Across all variables, MSM were the most likely to report less sex than in 2019 (56/80; 70.0%) and those in a cohabitating relationship were the most likely to report the same amount of sex (146/321; 45.5%), or more sex than in 2019 (62/321; 19.3%, Supplementary Table 1).

Sexual practices
When asked whether participants had changed their sexual practices because of COVID-19, 14.6% (123/840) reported that they were using sex toys more often on their own and 26.0% (218/838) reported that they were masturbating more. When stratified by frequency of sex during lockdown, those who reported less or no sex during lockdown were more likely to report using sex toys alone (18.3% vs 8.3%, diff=10.0%; 95%CI: 5.5, 14.6) and masturbating alone (35.6% vs 10.3%; diff=25.3%; 95%CI: 20.0, 30.6) compared with those who reported the same amount or more sex since COVID-19 ( Figure 1). A total of 98 participants (11.5%) reported buying a sex toy during lockdown and of these, 24.0% indicated that this was their first.
. CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 12, 2020. . https://doi.org/10.1101/2020.08.10.20171348 doi: medRxiv preprint  *gender diverse includes trans and non-binary; †the difference in proportions lockdown-2019; calculated from logistic regression model; ‡MSM=men who have sex with men (including bisexual men); WSW=women who have sex with women (including bisexual women); Other=asexual and non-binary, and those who selected 'something else'; § 'In a relationship' includes those who selected married/de facto/boyfriend/girlfriend/LAT (living apart together) or long distance relationship; 'cohabiting' defined as living with partner/s; Single includes those who selected single/divorced/widowed (and did not select another relationship); Other includes people who selected polyamorous or multiple partners (and did not select another relationship option, such as single), and those who indicated in comments that they were casually dating, but not exclusive.  Figure  2). The change in app use between time periods did not vary by gender but varied significantly by age for swapping pictures (diff =-9.2% for 18-29 years versus 4.9% for 30+; p=0.02) and by sexuality for hook-ups (diff=-63.9% MSM versus -22.9% for heterosexuals; p<0.01). No other differences were found (Supplementary Table 3).

DISCUSSION
Findings presented here demonstrate clear changes in sexual activity and sexual practices during the peak of the COVID-19 pandemic (to-date) in Australia and the lockdown measures it prompted. Our findings show a decline in sexual activity during the lockdown period with more than half of participants (53.5%) reporting having less sex during lockdown as compared to 2019. These patterns were most stark among those who reported being single, with 69.1% reporting less sex compared to those who reported being in a relationship. Somewhat unsurprisingly, of those reporting sexual activity during lockdown, sexual partners were most often regular partners, with few reporting sexual activity with casual hook-ups (7.8%). While research on the impact of COVID-19 on sexual behaviour is currently limited, data from a cross-sectional study conducted in the United Kingdom among those self-isolating reported similar results, with participants who were married or in a domestic relationship more likely to report sexual activity in the past week than their single counterparts [17].
Alongside declines in sexual activity, our findings also demonstrate changes in sexual practices. As could be anticipated given reduced opportunity for meeting partners, our findings show an increase in solo sex activities, including masturbation and using sex toys, particularly among those reporting less or no sex during lockdown. Indeed, adult stores in Australia reported a surge in sales during lockdown [18]. We were also interested in whether participants had changed their normal sexual practices or were engaging in additional hygiene practices, like washing their hands before and after sex, due to COVID-19. Although some participants indicated more frequently washing their hands before and after sex, over all we saw little change in partnered sex practices.
Contrary to our initial assumptions that dating app use would increase during lockdown, overall our findings showed a marked decrease in use. In particular, we saw significant declines in use among people who identified as female and aged 18-29 years. Dating apps are often used to facilitate in-person sexual and romantic connections [19], and the physical distancing enforced during lockdown perhaps drove usual users off the platform. Of those still using apps during . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 12, 2020. . lockdown, while some continued to organise in-person dates/hook-ups, we saw a significant increase in use for chatting/texting and organising virtual dates. Interestingly, rates of sexting or swapping intimate pictures did not significantly change between 2019 and lockdown. Further, only a small proportion of participants reported being diagnosed with STIs during lockdown and few reported accessing STI testing. However, given that some participants reported arranging in-person dates/hook-ups during lockdown, ready access to STI screening and treatment services during the COVID-19 pandemic is vital.
Our findings should be interpreted within their limitations. Namely, we used convenience sampling to recruit participants, and our resultant sample was largely homogenous, with most participants identifying as female, aged <30 years, and well educated. We also had an overrepresentation of WSW. Further limitations include recall bias, particularly for data on activity during 2019, and missing data for several variables (although this was usually ≤13%). Several participants also initiated the survey but did not complete it. Those who completed the survey were more likely to be older than those who did not. However, our study is novel in providing unique insight into changes in sexual activity during the peak of COVID-19 lockdown restrictions (to-date) among a cohort of people living in Australia and its strengths will be realised in subsequent cohort analyses of future waves of data.
The COVID-19 pandemic and the lockdown measures it prompted clearly impacted the sexual activity of people living in Australia. Although restrictions are beginning to ease across most of the country, recent spikes in cases in Melbourne [20], the second largest city in Australia, serve as a reminder that the pandemic is still with us and probably will be for some time. It is therefore essential to continue to monitor changes in sexual activity, and associated implications for sexual and reproductive health. In the short term, as restrictions lift and people increasingly engage in casual sex, sexual health organisations have produced guidelines for reducing risk of COVID-19 transmission during these encounters, and are encouraging regular HIV and STI screening [13]. Others warn of the continued impact of the pandemic on sexual and reproductive health, including reduced access to abortion services and an increase in intimate partner violence [21]. Whether or not Australia will experience an increase in fertility, as has been observed after high-mortality disasters like the 2004 Indonesian tsunami [22], or a long-term fertility reduction as seen in Sweden after the 1918 pandemic [23] is yet to be seen.

Acknowledgments
We would like to thank everyone who generously gave their time to complete our survey.

Competing interests
EC reports grants from the National Health and Medical Research Council, outside the submitted work. JH is supported by a National Health and Medical Research Council Fellowship (1136117). The other authors report no competing interests.

Funding
This study did not receive any external funding. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 12, 2020. . https://doi.org/10.1101/2020.08.10.20171348 doi: medRxiv preprint All authors contributed to the design and development of the survey. JC was responsible for administering the survey. HB and JH conducted the analysis. JC, HB, JH and FK interpreted the results and drafted the manuscript. All authors contributed to the revision of draft iterations of the manuscript prior to submission.   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 12, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 12, 2020. * Chi Square p value; † gender diverse includes trans and non-binary; ‡ MSM=men who have sex with men (including bisexual men); WSW=women who have sex with women (including bisexual women); Other=asexual and non-binary, and those who selected 'something else'; §'In a relationship'= married/de facto/boyfriend/girlfriend/LAT (living apart together) or long distance relationship; 'cohabiting'=living with partner/s; Single=single/divorced/widowed (and did not select another relationship); Other=polyamorous or multiple partners (and did not select another relationship option, such as single), and those who indicated in comments that they were casually dating, but not exclusive.   Diff= difference in proportions, 2019-lockdown. Calculated from logistic regression model ; *N=number of people reporting sex during the time period and answering the question; † gender diverse includes trans and nonbinary; ‡ MSM=men who have sex with men (including bisexual men), WSW=women who have sex with women (including bisexual women), Other=asexual and non-binary, and those who selected 'something else'; §'In a relationship'= married/de facto/boyfriend/girlfriend/LAT (living apart together) or long distance relationship; 'cohabiting'=living with partner/s; Single=single/divorced/widowed (and did not select another relationship); Other=polyamorous or multiple partners (and did not select another relationship option, such as single), and those who indicated in comments that they were casually dating, but not exclusive n/N* (%) -11.5 (-33.0, 10.0) Diff = difference in proportions, lockdown-2019. Calculated from logistic regression model; *N=number of people reporting sex during the time period and answering the question; †gender diverse includes trans and non-binary; ‡MSM=men who have sex with men (including bisexual men); WSW=women who have sex with women (including bisexual women); Other=asexual and non-binary, and those who selected 'something else' §'In a relationship'= married/de facto/boyfriend/girlfriend/LAT (living apart together) or long distance relationship; 'cohabiting'=living with partner/s; Single=single/divorced/widowed (and did not select another relationship); Other=polyamorous or multiple partners (and did not select another relationship option, such as single), and those who indicated in comments that they were casually dating, but not exclusive.

Introduction
Background/rationale 2 Explain the scientific background and rationale for the investigation being reported 3 Objectives 3 State specific objectives, including any prespecified hypotheses 3

Study design 4
Present key elements of study design early in the paper 4  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 12, 2020. . https://doi.org/10.1101/2020.08.10.20171348 doi: medRxiv preprint (b) Report category boundaries when continuous variables were Discuss the generalisability (external validity) of the study results 10-11

Funding 22
Give the source of funding and the role of the funders for the present study and, if applicable, for the original study on which the present article is based 11 *Give information separately for exposed and unexposed groups. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 12, 2020. . https://doi.org/10.1101/2020.08.10.20171348 doi: medRxiv preprint