Supporting our first two predictions, greater social connectedness in the last seven days and greater nature exposure (both in terms of nearby nature, and visits in the last week) were both associated with greater subjective wellbeing (SWB) over the same period. Moreover, in partial support of our key third prediction, nearby nature, but not visit frequency, moderated the relationship between social connectedness and SWB over the last seven days. Put simply, a relative lack of social connectedness was less important for SWB among those who lived in greener areas. Or to put it another way, nearby nature appeared to ‘buffer’ the effect of a lack of social connectedness on wellbeing, supporting and extending previous work that also suggested that living in more natural areas buffers against the adverse impacts of negative events and circumstances on stress and wellbeing (e.g., [94
]). Importantly, this effect was replicated across both the overall WHO-5 score and a binary version which identifies a cut-off for being at risk of depression [104
]. That is, high levels of nearby nature appeared to reduce the likelihood of showing signs of depression among individuals with low social connectedness. Contrary to predictions, however, intentional exposure, i.e., the number of visits made to nature in the last week, did not show a similar buffering effect, suggesting instead that socially connected individuals stand to gain roughly as much as socially isolated individuals from making more frequent visits to natural areas in terms of their SWB.
That the strength of association between our measures of nature exposure and SWB were similar in size to those between social connectedness and SWB shows just how important the local environment can be, given that social connectedness is often thought to be one of the key predictors of wellbeing [10
]. The buffering effect of nearby nature on the social connectedness-SWB relationship also has important implications for the domain-based conceptualisation of human wellbeing, because it would appear that a deficit in the “social domain” of SWB may be supplanted by experiences based in the biophysical environment. This “inter-domain” interaction thus supports the possibility that both social connectedness and nature exposure may act upon shared mechanistic pathways (e.g., stress, rumination, self-esteem, etc.), and that they may be expressions of the same underlying need, i.e., the need to feel connected, whether this is manifested as being connected to other people or the natural world. Although from an evolutionary perspective this may suggest that both social and environmental factors are indicative of the perceiver’s reproductive fitness, we remain cautious about proposing evolutionary explanations given the more easily demonstrable influences of culture and individual experiences [120
]. Clearly far more, and possibly experimental, work is needed to unpick these possibilities.
There are a number of possible reasons why we did not see an interactive effect between social connectedness and visit frequency. One possibility is due to a restricted data range. Specifically, although nearby nature had a good distribution across the 1–7 scale, approximately a third of participants reported not visiting the natural environment at all in the last seven days and only 25% reported visiting more than once or twice. Thus although, in some ways, nature visit frequency is a more directly comparable measure of ‘contact’ to the frequency of social contact experiences, it has a more restricted range than nearby nature, which people may still experience indirectly (e.g., through window views, e.g., [11
]). It may also be that visiting nature is not as “clean” a measure of nature exposure as nearby nature in that it also (usually) involves other things which promote wellbeing but which were not included in the current analysis such as some form of physical activity [15
]. Although we are unsure exactly why the associations between the measures differed here, we nonetheless believe it is important to test both, and other, operationalisations of nature exposure in future work to help us gain a better understanding of how different measures of nature exposure reflect different types of experience.
A relatively novel aspect of the current work was its inclusion of personality among the set of covariates. This was important because it meant that an aspect of an individual that might influence both their desire for contact with others, and their desire for contact with nature, was controlled for, reducing the possibility that any interaction was due to an unknown third factor (in this case personality). Reflecting earlier work [4
], extraversion was positively related to SWB, and importantly for current purposes it was also significantly positively related to both measures of nature exposure, although somewhat surprisingly was non-significantly positively related to social connectedness (see Supplementary Table S1
). In other words, by including extraversion in our covariate set we were able to reduce the potential problem that extraverts are both happier and more likely to report higher nature exposure, and thus any association between SWB and nature exposure may simply be due to their shared variance with extraversion. We suggest that future studies exploring nature exposure and health and wellbeing outcomes may also want to attempt to control for this often unobserved yet stable source of heterogeneity.
Of course, we also recognise several limitations with the current study. First, the sampling methods could be improved. Due to restrictions on budget, time and resources associated with the current MSc project, the sample size was small and was not weighted to be population representative. As such, the results are not yet generalizable beyond the current sample. Future research might therefore want to explore whether the current findings might replicate across larger, more population representative surveys such as the Monitor for Engagement with the Natural Environment [108
]. Second, even though personality was included, the cross-sectional methodology still prohibits causal inference, although it would be hard to explain the interactive effect between social connectedness and nearby nature though a reverse direction account from wellbeing to the moderating effect of nature exposure on social contact. Third, all variables were subject to the limitations of self-report data, and we were unfortunately unable to collect objective nearby nature in the current study. Mitigating this issue however, is the suggestion that subjective measurements of nature exposure may be of equal or greater importance than “objective” ones in any case [124
]. Nevertheless, further longitudinal survey data (e.g., [52
]) would enable future research to explore how fluctuations in social connectedness and nature exposure for the same individuals are related to fluctuations in mental health and SWB over time and experimental studies, perhaps similar to those by Hartig and colleagues [87
], would enable a more targeted exploration of our core hypothesis. Finally, we also recognise that several potentially important covariates were not included in the survey, including employment status, physical health, and physical activity level, which can all have strong influences on mental wellbeing [6
]. As noted above the lack of physical activity data, and thus the ability to control for it, may in part explain our inability to find a moderating effect of nature visit frequency on the social connectedness-SWB relationship.
Importantly, we wish to stress that the ways we operationalised SWB, social connectedness and nature exposure may not reflect individuals’ entire experiences of the phenomena we wished to measure. In particular, because participants only reported on their experiences in the seven days before they took the survey, it would be unwise to suggest that our results generalise to related phenomena in the longer-term, such as extended periods of social isolation.
These limitations notwithstanding, results have potential implications for policy and practice in dealing with the issues of social isolation and population wellbeing. Specifically, our data suggest that the greening or naturalisation of residential environments may provide a buffer against social isolation, in a similar way to other stressors. In the UK, for instance, those most at risk of social isolation tend to be older individuals, ethnic minorities, and those living in more deprived areas; the exact same groups who tend to have lower than average contact with nature [98
]. Consequently, increasing the quantity of nature in neighbourhoods dominated by these and other isolated population segments is a potentially important public health good. At the individual-level, our results suggest that people who are socially disconnected, for instance when moving to a new area, may avoid the worst effects of social isolation if they move to a more natural location. Finally, our results suggest that nature visit frequency may be beneficial for SWB at various levels of social connectedness, and so even the most socially connected person may still stand to benefit from actively going out to experience nature.