- 5.0% of people in Great Britain (2.6 million adults) reported that they felt lonely "often" or "always" between 3 April and 3 May 2020, about the same proportion as pre-lockdown.
- Of those asked, 30.9% (7.4 million adults) reported their well-being had been affected through their feeling lonely in the past seven days.
- Working-age adults living alone were more likely to report loneliness both “often or always” and over the past seven days than the average adult; this was also the case for those in "bad" or "very bad" health, in rented accommodation, or who were either single, or divorced, separated or a former or separated civil partner.
- Both those feeling lonely “often or always” and in the past seven days had lower personal well-being scores including higher anxiety scores than the Great Britain average, but the effect was stronger among those feeling lonely “often or always”.
- Both those feeling lonely “often or always” and in the past seven days were more likely than the average to say they were struggling to find things that help them cope during lockdown.
- Around 7 in 10 of those feeling lonely “often or always” “agreed” or “strongly agreed” that they had people who would be there for them, compared with 9 in 10 of the Great Britain average.
“The Office for National Statistics has been researching people’s well-being for nearly a decade, providing a different perspective on how our country is doing, and on social inequalities. "Lockdown" affected everyone, but responses differed. During that first month, the equivalent of 7.4 million people said their well-being was affected through feeling lonely.
"Lonely" people were more likely than others to be struggling to find things to help them cope and were also less likely to feel they had support networks to fall back on.”
Dawn Snape, Assistant Director of Sustainability and inequalities Division, ONSNôl i'r tabl cynnwys
We have looked at two measures of loneliness as part of this release using data from the Opinions and Lifestyle Survey. For more information, see “Measures of loneliness” within the Measuring the data section.
- "Chronic loneliness” – this measures the percentage of those who feel lonely “often or always”; this question was asked to all respondents (5,260 adults).
- “Lockdown loneliness” – this measures the percentage of those who said their well-being had been affected through having felt lonely in the last seven days; this question was only asked to respondents who had already reported that their well-being had been affected in the past seven days and that they were “very” or “somewhat worried” about the effect of the coronavirus on their life (2,440 adults).
From 3 April to 3 May 2020, 5.0% of people said they feel lonely “often or always”, around 2.6 million people across Great Britain. Throughout this release we refer to this group as “chronically lonely”.
During the same period, 30.9% of those asked said their well-being had been affected through feeling lonely in the past seven days, around 7.4 million people across Great Britain. We refer to this group as “lockdown lonely”. If we adjusted this to be representative of the entire Great British population this would be equivalent to 14.3%.
We investigated measures of loneliness from a range of sources to see whether people are more or less lonely since the lockdown began, and who is most affected. We found that while reported estimates vary between surveys because of differences in wording or in how the survey was completed, chronic loneliness does not seem to have changed significantly as a result of lockdown.
Using data from the Opinions and Lifestyle Survey, we have also been able to explore the circumstances of people who report loneliness during the last week. There are fewer studies to use as comparators for this measure so it is not clear how it may have evolved at the Great Britain level relative to any given week prior to the lockdown. See Measuring the data.
How are our two loneliness measures linked?
To understand the interaction and overlap between the two groups, we took the lockdown loneliness question as our reference point and broke down the different combinations of chronic and lockdown loneliness among those who were asked both questions, giving a total of six groups. Results are summarised in Table 1, along with estimates of the numbers of adults who fall into each group.
Notably we saw that 6.3% of respondents suffered from both chronic and lockdown loneliness, while 1.3% suffered from lockdown loneliness despite saying that they “hardly ever or never” feel lonely. A further 1.6% of those asked said they had not felt lonely in the past seven days, despite feeling lonely “often” or “always”.
One possible explanation for this from our analysis could be that those who are chronically lonely are more likely to have an underlying health condition or disability, and therefore be more likely to be vulnerable to the coronavirus and more strictly social distancing as a consequence (see “Demographics of loneliness”). It is possible that this group may be receiving more support during the lockdown than they would have otherwise had, or that they are able to engage in a wider range of activities now offered virtually than was previously possible.
|Felt lonely in last seven days||Hasn't felt lonely in past seven days|
|"Often" or "Always" lonely||6.3% (1.5 million adults)||1.6% (0.4 million adults)|
|Both chronically lonely and lockdown lonely||Chronically lonely who are not lockdown lonely¹|
|"Some of the time" or "Occasionally"||23.5% (5.6 million adults)||31% (7.4 million adults)|
|"Hardly ever" or "Never" lonely||1.3% (0.3 million adults)||36.3% (8.6 million adults)|
|Lockdown lonely but not chronically lonely||Neither lockdown nor chronically lonely|
Download this table.xlsx .csv
Demographics of loneliness
We found that for both measures, factors linked to loneliness fell under the broad categories of health or household circumstances. Many of these characteristics reflect the findings of previous loneliness research undertaken by the Office for National Statistics (ONS).
|Chronic loneliness||Lockdown loneliness|
|All below characteristics are significantly different relative to the Great Britain population figure of 5.0%||All below characteristics are significantly different relative to the Great Britain population figure of 30.9%|
|More likely to be lonely||More likely to be lonely|
|· Reported "bad" or "very bad" (26.3%)||· Reported "bad" or "very bad" (52.1%)|
|· Disabled (11.0%)||· Single (53.0%)|
|· Reported underlying health condition (8.8%)||· Widowed (51.2%)|
|· Single (9.0%)||· Divorced or separated/former or separated civil partner (53.2%)|
|· Divorced or separated/former or separated civil partner (9.7%)||· Adults living alone (aged 16 to 64 years) (57.2%)|
|· Adults living alone (aged 16 to 64 years) (13.4%)||· Adults living alone (aged 65 years and over) (50.1%)|
|· Rented accommodation (8.6%)||· Rented accommodation (38.8%)|
|· Aged 16 to 24 years (50.8%)|
|Less likely to be lonely||Less likely to be lonely|
|· Married/cohabiting/same-sex cohabitee/civil partner (2.5%)||· Married/cohabiting/same-sex cohabitee/civil partner (16.5%)|
|· Two-adult households (at least one aged 65 years and over) (1.7%)||· Two-adult households (at least one aged 65 years and over) (14.3%)|
|· Own home outright (3.1%)||· Aged 55-69 years (24.1%)|
|Characteristics with no significant difference||Characteristics with no significant difference|
|· Great Britain country||· Great Britain country|
|· Sex||· Sex|
|· Having children||· Disability|
|· Giving special help to a sick, elderly or disabled adult in the household||· Reported underlying health condition|
|· Having children|
|· Giving special help to a sick, elderly or disabled adult in the household|
Download this table.xlsx .csv
For both groups, people who reported "bad" or "very bad" health were more likely than the Great Britain average to be lonely. However, there was no evidence of a link between lockdown loneliness and more objective indicators of health or wellness, such as having an underlying health condition or disability.
This suggests that ongoing health concerns may be more strongly associated with chronic loneliness. Further research on loneliness and disability can be found in our release on Coronavirus and the social impacts on disabled people in Great Britain, and on Disability, well-being and loneliness.
People who were married, cohabiting or in a civil partnership (both same and opposite sex) were less likely than the Great Britain average to report either chronic or lockdown loneliness, while those who were either single, or divorced, separated, or a former or separated civil partner were more likely.
Household composition shows a similar pattern, with those in older two-adult households (where at least one person is over the age of 65 years) being less likely to report either chronic or lockdown loneliness than the Great Britain average, while working-age adults living alone were more likely. Living in rented accommodation was also linked to increased reporting of loneliness under both measures.
Finally, we found that those aged 16 to 24 years were more likely to have experienced lockdown loneliness (50.8%), while those aged 55 to 69 years were less likely (24.1%) to have done so. This echoes previous research exploring chronic loneliness showing that people in younger age groups were most likely to report loneliness, while those in older age groups were less likely. However, we note that our oldest age band, those aged 70 years and over, were no less likely than average to report lockdown loneliness.
We have not tested here whether people with multiple characteristics associated with loneliness are at greater risk of experiencing loneliness. However, we have previously published a latent class analysis, which explores how different characteristics may be linked.
Lockdown loneliness: which factors tip the balance for the "sometimes lonely"?
Feelings of loneliness affect most of us at some point in our lives, and indeed, 43.4% of all respondents said that they felt lonely either "some of the time" or "occasionally". But is there any way of differentiating which individuals within this category are more likely to have experienced lockdown loneliness?
For those respondents feeling lonely "some of the time" or "occasionally" that had also been asked the lockdown loneliness question, we looked at the demographics of those who experienced lockdown loneliness, and those who did not. Our findings, summarised in Table 3, suggest that younger people and those living alone are at greater risk of lockdown loneliness.
|Lonely "some of the time" or "occasionally"||Lonely "some of the time" or "occasionally"|
|More likely to experience lockdown loneliness||Less likely to experience lockdown loneliness|
|· Aged 16 to 39 years||· Aged 40 to 69 years|
|· Single, or divorced, separated or a former or separated civil partner||· Married, cohabiting, same-sex cohabitee or civil partner|
|· Living in a single-person household||· Living in a multi-person household|
Download this table.xlsx .csv
We also found that those in this group who did experience lockdown loneliness had lower average happiness scores compared with those who didn’t experience lockdown loneliness. However, there was no significant link to any of the other personal well-being questions (life satisfaction, feeling the things they do in life are worthwhile, and anxiety).Nôl i'r tabl cynnwys
While our two lonely groups have shown broadly similar experiences in how their well-being has been impacted by the coronavirus (COVID-19), we also saw that they were both more likely to report that they were struggling to find things to help them cope while staying at home.
Looking at areas that were helping people cope, we do see that people feeling lonely have not necessarily responded to lockdown in the same way.
Individuals in the chronically lonely group were no more likely to use any particular activity to help them cope than the Great Britain average. However, they were less likely than average to use the following activities to help them cope:
spending time with other household members
keeping up with family and friends
Those in the lockdown lonely group were also less likely to spend time with other household members or gardening to help them cope; but they were more likely to watch films or other online sources of entertainment such as museum tours or theatre productions.
The reduced take-up in both lonely groups for spending time with other household members and gardening reflects the different living environments of those we found to be more likely to report loneliness, such as being more likely to live alone or to live in rented accommodation. We also note that people living alone are more likely to be renting their homes, and that rented properties are less likely to have a garden at all, with 24% of those in rented homes in England saying they had no access to a garden, compared with 4% of those in owned homes (source: Monitor of Engagement with the Natural Environment, 2014 to 2019. See Measuring the data for more information).
For many of us, lockdown has meant that our interaction with our usual support networks has changed, and for some this has meant turning to the local community for support when we have needed it.
When asked whether they felt there were people in the community who could support them, less than half of the chronically lonely (45%) said they "agreed" or "strongly agreed". A further 25% said they "did not know". Among the lockdown lonely, 56.7% "agreed or strongly agreed" that they could get support from the community if they needed it. For both groups, this sense of community support was lower than the Great Britain average, which was closer to two-thirds (66.3%).
When asked if they had people who would be there for them if they needed help, both groups were also less likely to say they "agreed" or "strongly agreed" than the Great Britain average. Figure 6 shows that just under 7 in 10 (69.4%) of the chronically lonely gave this response, over 20 percentage points less than the Great Britain figure (92.3%).
Figure 6: Around 7 in 10 of those feeling lonely “often or always” "agreed" or "strongly agreed" that they had people who would be there for them
Percentage of people who said they had people who would be there for them, Great Britain, 3 April to 3 May
All respondents were asked "How often do you feel lonely?". Therefore the base population for those reporting "often or always" lonely is those aged 16 years and over in Great Britain.
Respondents were asked "In the past seven days, how has your well-being been affected?" giving the option for noting "Feeling lonely" if they have already noted that they were "very worried" or "somewhat worried" about the "effect that Coronavirus (COVID-19) is having on their life right now?" and if they reported "My well-being is being affected (for example, boredom, loneliness, anxiety and stress)" when asked "In which ways is Coronavirus (COVID-19) affecting your life?".
"*" indicates where the sample size for that response is less than 26 and caution should be applied.
This release provides the first in-depth look at loneliness during lockdown. Later this month, we will soon be publishing more detailed analysis of the impact of different characteristics and circumstances on anxiety during lockdown, including loneliness.
We will also continue to monitor changes in loneliness as lockdown measures are eased.Nôl i'r tabl cynnwys
To define disability in this publication, we refer to the Government Statistical Service (GSS) harmonised "core" definition: this identifies as "disabled" a person who has a physical or mental health condition or illness that has lasted or is expected to last 12 months or more that reduces their ability to carry-out day-to-day activities.
The GSS harmonised questions are asked of the respondent in the survey, meaning that disability status is self-reported.
In this release, our definition of loneliness is aligned to the definition used by the Jo Cox Commission and in the Loneliness Strategy for England. This is based on a definition first suggested by Perlman and Peplau in 1981: "A subjective, unwelcome feeling of lack or loss of companionship. It happens when we have a mismatch between the quantity and quality of social relationships that we have, and those that we want."
We have used two measures of loneliness in this release. Please see "Measures of loneliness" in Measuring the data for more information.
Underlying health condition
In this analysis, underlying health conditions were defined based on NHS guidance around those at higher risk of contracting the coronavirus. This guidance changed over the period covered within this release.
Within our analysis, adults with an underlying health condition include those with:
angina or long-term heart problem
autism spectrum disorder (ASD) or Asperger's (Asperger syndrome)
chronic obstructive pulmonary disease (COPD) or long-term lung problem
kidney or liver disease
From 9 April this also includes pregnant women.
From 24 April this also includes:
conditions affecting the brain and nerves, such as Parkinson's disease
a weakened immune system as the result of conditions such as HIV and AIDS, or medicines such as steroid tablets or treatment for cancer
problems with your spleen - for example, sickle cell disease, or if you've had your spleen removed
being overweight (having a BMI of 40 or above)
given an organ transplant
From 24 April this excludes:
Alzheimer's disease or dementia
epilepsy or other conditions that affect the brain
high blood pressure
stroke or cerebral haemorrhage or cerebral thrombosis
Lockdown is the shutting down of all non-essential activities to slow the spread of the coronavirus (COVID-19). In the UK, this has seen strict limits imposed on daily life, including:
people ordered to only leave the house for essentials such as food, medicine, exercise or to care for a vulnerable person
the closure of non-essential shops
the banning of gatherings of more than two people
The UK lockdown was applied on 23 March 2020. This has formed the basis for each nation's stay-at-home guidance. Specific stay at home guidance for England, Scotland, Wales and Northern Ireland is available. During the period covered in this release, all countries of the UK remained in lockdown but the data in this release only cover England, Scotland and Wales.Nôl i'r tabl cynnwys
More quality and methodology information on strengths, limitations, appropriate uses, and how the data were created is available in the Opinions and Lifestyle Survey QMI.
Measures of loneliness
We have used two measures of loneliness in this release: feeling lonely "often" or "always" (referred to as chronic loneliness in this release) and feeling lonely in the past seven days (referred to in this release as lockdown loneliness).
Feeling lonely "often" or "always"
This measure is taken from responses to the question: "How often do you feel lonely?" where response categories are:
often or always
some of the time
The question was asked to all respondents on the Opinions and Lifestyle Survey (OPN). Our analysis focuses on those responding "often or always" to this question; this constituted approximately 250 responses. The relatively small sample size of this group means that confidence intervals may be wider than normally anticipated, and some breakdowns of this group, such as the areas people are most worried about, require additional caution in interpreting.
These responses may be from people who were feeling lonely before the coronavirus pandemic for whom a period of enforced social isolation through lockdown may have prompted their feelings of loneliness to become more frequent.
Cognitive interviewing suggests respondents may be more likely to answer openly in self-completion formats (paper and online modes). As some interviews on the OPN were conducted via telephone, it is possible that there may be some underreporting of loneliness in this data source.
The Office for National Statistics regularly monitors loneliness as part of their Social capital and Measuring National Well-being releases. This uses the measure of feeling lonely "often" or "always", as measured by the Community Life Survey.
Feeling lonely in the past seven days
This measure considers those people who indicated "yes" to the option "Feeling lonely" within the multiple-choice question "In the past seven days, how has your well-being been affected?".
In the Opinions and Lifestyle Survey, the question was asked to those who had indicated that they were "very worried" or "somewhat worried" about the effect that the coronavirus (COVID-19) is having on their lives, and had subsequently indicated "yes" to the option "My well-being has been affected", when asked the follow-up multiple choice question "In which ways is coronavirus (COVID-19) affecting your life?". This constituted approximately 750 responses indicating that the respondent had felt lonely within the past seven days.
It is possible that the routing of this question has led to underreporting of those who have felt lonely in the past seven days. This is because people who had felt lonely may not have had the opportunity to answer the question.
This question is not included within the harmonised definition of loneliness; however, the Scottish Government include a similar measure of having felt lonely in the past seven days on the Scottish Household Survey. While loneliness is measured in Wales, differences in the wording of the question mean it has not been possible to compare our data with figures for Wales in this instance. See "Loneliness in the devolved administrations" for more information.
When scoping the value of using two measures of loneliness, we found evidence that our two measures of loneliness are linked, with four out of five people who are chronically lonely also reporting that they are lockdown lonely (79.8%).
However, that is not to say that the two questions are measuring the same experiences: with around one in five (20.3%) of our lockdown lonely group saying they were also chronically lonely, we also note that the majority of individuals within this group would not normally fall under the remit of our typical loneliness studies. This tells us that people's experiences of loneliness during lockdown are more varied than is typically considered.
Comparing measures over time
To understand how chronic loneliness may have changed since the lockdown, we looked at sources including the Opinions and Lifestyle Survey, the Community Life Survey, the Active Lives Survey, and Understanding Society. While many of these sources were not directly comparable with our data because of differences in factors such as question wording, geography, or how the survey was completed, the overall picture from looking at these sources together suggested that chronic loneliness does not seem to have changed significantly since lockdown.
There are fewer studies to use as comparators for our lockdown loneliness measure so it is not clear how it may have evolved at the Great Britain level relative to any given week prior to the lockdown. While we also looked at pre-lockdown figures of loneliness in the past week from the Scottish Household Study and the European Social Survey, differences such as geography, question wording and base sample meant we could not draw direct comparisons.
Pooling data from the Opinions and Lifestyle Survey
Statistics in this release have been taken from four waves of the Opinions and Lifestyle Survey (OPN). The OPN is a monthly omnibus survey. In response to the coronavirus (COVID-19) pandemic, we adapted the OPN to become a weekly survey used to collect data on the impact of the coronavirus pandemic on day-to-day life in Great Britain.
To enable more detailed analysis, such as the loneliness breakdowns presented in this release, four waves of this weekly OPN data have been pooled together and reweighted to create a larger dataset. By pooling data, we have improved the sample size available to create smaller breakdowns of individual questions at the expense of having to report on a wider time period (one month rather than one week).
This pooled dataset contains 5,320 individual responses, representing an overall response rate of 66% for the waves of the survey conducted from 3 April to 3 May 2020, meaning the entire reference period took place during lockdown. Survey responses were collected using an online self-completion questionnaire, with the option to take part over the phone.
The survey results are weighted to be a representative sample for the population. Weights were first adjusted for non-response and attrition, then calibrated to satisfy population distributions considering the following factors: sex by age, country or region, tenure, highest qualification, employment status, National Statistics Socio-economic Classification (NS-SEC) group and smoking status.
For age, sex and geography, population totals based on projections of mid-year population estimates for April 2020 were used. The resulting weighted sample is therefore representative of the Great Britain adult population by a number of socio-demographic factors and geography.
Some survey questions asked for people's responses in reference to "the past seven days". These results have been presented representing people's views during the month of April, even though people's attitudes may have changed slightly over our reference period.
Variables adapted during pooling
When creating a pooled dataset, it is only possible to include questions that were included across the entire period in question. Therefore, where additional response options have been added to multiple choice questions, we have incorporated them into existing categories so that people's responses could still be included in our analysis. Within this release, the following measures were affected. Full details may be found in thedatasets.
COVID-19 affecting life
From 24 April 2020, an additional response option was added to the multiple-choice question: "In which ways is Coronavirus (COVID-19) affecting your life?". This was "My exercise routine is being affected". Responses within this category were incorporated into responses listed as "other".
For consistency, the same approach was taken to the question "Of the worries you told us about, which one are you most concerned about?".
From 9 April 2020, two additional response options were added to the multiple-choice question: "In the past seven days, how has your well-being been affected?". These were "finding working from home difficult" and "unable to exercise as normal". Responses within each of these categories were incorporated into responses listed as "other".
Four additional response options were added to the multiple-choice question: "What is helping you to cope whilst staying at home?" on each of 9 and 17 April 2020. Responses to this question were subsequently regrouped as follows:
spending time with others that you live with
staying in touch with family and friends remotely
exercise (indoors and outdoors)
learning something new
watching films, arts and culture (for example, virtual museum tours and online plays)
online activities, support and information (including using the internet to continue your usual activities (for example, virtual worship services))
other, including limiting how often you listen to or watch the news, following lockdown tips and guides made by celebrities and athletes, struggling to find things that are helping me cope, arts activities, DIY activities
It is also known that for interviews conducted between 17 April 2020 and 26 April 2020, this question was not asked to those who responded via telephone. These respondents were therefore removed from analysis of this question.
Assessment of differences
The measures analysed in this release are sourced from self-reported survey data. These sources use samples of the total measured population to produce estimations. Given this, measures for given groups have only been assessed as higher or lower than the comparison group if the difference between groups is statistically significant using 95% confidence intervals. If a difference is said to be statistically significant, it is unlikely that it could have occurred by chance.
When looking at the demographics of our two lonely groups in Table 2, we break our sample down into five age bands: 16 to 24 years, 25 to 39 years, 40 to 54 years, 55 to 69 years and 70 years and over.
However, because of the smaller sample size, when analysing demographics of the "sometimes" or "occasionally" lonely group in Table 3, we collapse these categories, so analysis is provided by three age bands: 16 to 39 years, 40 to 69 years and 70 years and over. These broader age categories allow us to provide more robust analysis of this group.
Access to gardens
In our section on coping under "How are lonely people responding?" we referred to data breaking down access to gardens by tenure. The data are summarised more fully in Table 4.
|Rent (both private|
and socially rented)
|Owned (both outright|
and with a mortgage)
|I don't have access to a garden||24%||4%|
|I have access to a private communal garden||8%||2%|
|I have access to a private garden||64%||91%|
|I have access to a private outdoor space |
but not a garden (balcony, yard, patio area)
Download this table.xlsx .csv
For more information on these data please see the Monitor of Engagement with the Natural Environment (MENE).
Loneliness in the devolved administrations
The UK's devolved administrations also collect and publish data on loneliness.
Scotland captures data on loneliness through the Scottish household survey, and its question on feeling lonely in the last week is included within the "Social networks" theme of Scotland's social capital index. This is a national indicator that forms part of Scotland's National Performance Framework. It monitors aggregate changes in levels of social capital since 2013 through the four domains of social networks, community cohesion, community empowerment and social participation. More information can be found on the National Performance Framework website.
In Wales, a measure of loneliness is included within the national indicators to demonstrate progress towards Wales' seven well-being goals. Further data relating to loneliness at a Wales level are collected through the National Survey for Wales.Nôl i'r tabl cynnwys
The main strengths of the Opinions and Lifestyle Survey (OPN) include:
It allows for timely production of data and statistics that can respond quickly to changing needs
It meets data needs: the questionnaire is developed with customer consultation, and design expertise is applied in the development stages
Robust methods are adopted for the survey's sampling and weighting strategies to limit the impact of bias
Quality assurance procedures are undertaken throughout the analysis stages to minimise the risk of error
The main limitations of the OPN include:
The sample size is relatively small: 2,010 individuals per week with fewer completed interviews, meaning that detailed analyses for subnational geographies and other sub-groups are not possible
The mode is online only, so the sample may be subject to more bias than usual
More quality and methodology information on strengths, limitations, appropriate uses, and how the data were created is available in Coronavirus and the social impacts on Great Britain and the Opinions and Lifestyle Survey QMI.Nôl i'r tabl cynnwys
Manylion cyswllt ar gyfer y Bwletin ystadegol
Ffôn: +44 (0)1633 651814