Can social support buffer the association between loneliness and hypertension? a cross-sectional study in rural China.

OBJECTIVES: Hypertension has reached epidemic levels in rural China, where loneliness has been a major problem among community dwellers as a consequence of rural-to-urban migration among younger generations. The objective of the study is to investigate the association between loneliness and hypertension, and whether social support can buffer the association (i.e., stress buffering theory), using cross-sectional data from 765 adults (mean age: 59.1 years) in rural Fujian, China.
METHODS: Social support was measured as the reciprocal instrumental social support from/to neighbors and the reciprocal emotional support (i.e., the number of close friends that the respondent could turn to for help immediately when they are in trouble). A mixed-effect Poisson regression model with a robust variance estimator was used to investigate the association between loneliness, social support, and hypertension.
RESULTS: Analysis revealed that those who were lonely had a higher prevalence ratio for hypertension (prevalence ratio = 1.12, 95% confidence interval 0.99-1.26) compared to those who reported not being lonely. There was an interaction between social support and loneliness in relation to hypertension. Specifically, contrary to the stress buffering theory, the positive association between loneliness and hypertension was more pronounced among those who reported higher social support compared to those who reported lower support (p for interaction <0.001 for instrumental support).
CONCLUSION: The results suggest that being lonely despite high levels of social support poses the greatest risk for hypertension. This study did not confirm a buffering effect of social support on the association between loneliness and hypertension.

Introduction

More than 1.13 billion people are affected with hypertension globally [1]. Among risk factors for hypertension (e.g., salt in the diet, overweight/obesity, tobacco use, and physical inactivity) [2], recent studies have suggested that loneliness, which has been defined as the subjective feeling that accompanies the perception that one’s social needs are not being met by the quantity or quality of one’s social relationships [3], may be an important risk factor for hypertension [4-6]. Chronic loneliness is considered as a psychological stressor, which could lead to worse health through behavioral choices (e.g., less physical activity, more daily smoking, and poor sleep), and could also be directly linked to impaired stress response (i.e., physiological functioning) [3]. Meta-analyses have shown that loneliness is a risk factor for coronary heart disease, stroke [7] and all-cause mortality [8].

In China, loneliness has been singled out as a major problem among rural community dwellers [9]. The combination of China’s one-child policy (from the years 1979 to 2015) and the massive rural-to-urban migration of working-aged adults has resulted in a high prevalence of socially isolated ‘left behind’ people, especially among older population [10]. According to surveys, the prevalence of loneliness among older people in rural areas ranges from 25% to 78% [9, 11]. These sociodemographic transitions might have underlain the increase in disease burden associated with hypertension; mortality attributable to hypertension in China almost doubled during the decade from 2007 to 2017 [12], with a larger increase observed in rural vs. urban areas [13, 14]. Despite this, to date, there have been no studies that examined the association between loneliness and hypertension in China.

In the absence of family support especially from adult children both mentally and instrumentally, social support in the community (i.e., the resources provided by one’s network with the intention of increasing one’s coping ability [15]) can be a key to mitigate the adverse effect of loneliness in rural Chinese society. Social support, which is commonly categorized into several types of behaviors (e.g., instrumental support and emotional support [16]), is suggested to be a coping resource to overcome loneliness [17] and affect health via a buffering mechanism [18]. Although the studies are not uniformly positive, some studies support a buffering effect of social support on the association between perceived stress and ambulatory blood pressure [19], as well as low income and diastolic blood pressure [20].

Against this background, the purpose of this study was to (1) investigate the association between loneliness and hypertension in rural Fujian, China, and to (2) investigate the potential buffering effect of social support in the association between loneliness and hypertension. The authors hypothesized that (1) loneliness is positively associated with hypertension; and (2) social support from neighbors would mitigate (buffer) the association between loneliness and hypertension (i.e., high social support would protect against the positive association of loneliness on hypertension).

Materials and methods

Field survey

A cross-sectional survey was conducted in a sample of seven rural communities in one city in Fujian Province, China in August 2015, selected on the basis of average population size and level of economic development. Median household income based on the responses from 161 household heads (i.e., 11,000 RMB, interquartile range: 4800–24000 RMB) was comparable to figure reported for rural communities in the area as of 2014 (i.e., 11,252 RMB) (1 RMB = 0.16 USD as of 2015) [21]. All residents aged 18 years or older (i.e., adults in China) were invited through an advertisement posted at each community health center, and about 62% of the population eventually participated (i.e., the convenience sampling). Questionnaire and anthropometric data were collected by trained staff from 797 participants. Details of the survey were described elsewhere [22].

Measures

Hypertension

Systolic and diastolic blood pressure (SBP and DBP) were measured in the sitting position with the left arm held horizontal at the level of the heart using an automated oscillometric monitor by the authors or staffs at local health centers (HEM-7000, OMRON Corp., Japan). They were measured twice and the mean values were calculated. Although there was no specific resting time before the measurement, people first registered and took instructions/explanations about the survey and then signed the certificate of consent before the measurement (~10 min/person). People were defined as having hypertension if they had SBP ≥140 mmHg or DBP ≥90 mmHg or if they were currently taking antihypertensive medication.

Loneliness

Loneliness was measured using the single item: ‘How often did you feel lonely in the past one month?’ with possible responses on a 5-point Likert scale. The answers were dichotomized: lonely (frequently and always, or sometimes) vs. not lonely (never, rarely). The single-item self-report measure has been shown to be highly correlated with an established loneliness scale (the University of California Los Angeles (UCLA) Loneliness Scale) comprising 20 items (e.g., r = 0.72) [23], and is widely used in field surveys [24].

Social support

Social support was measured in two dimensions: reciprocal instrumental support and reciprocal emotional support. Reciprocal instrumental support was measured as the reported frequency of exchanging (receiving or providing) various commodities (e.g., food and medicines) between neighbors. Data on the frequency of receiving and providing instrumental support were separately obtained to create four categories (i.e., neither receiving or providing instrumental support; only receiving instrumental support; only providing instrumental support; both receiving and providing instrumental support). Reciprocal emotional support was assessed as the number of close friends that the respondent could turn to for help immediately when they are in trouble, which was then categorized into three groups (none; 1 to 5; 6 or more) since the number of friends ranged from 0–35, and 48% of the participants answered they have no friend.

Demographic, socio-economic, and lifestyle factors

Demographic and socio-economic data included age (in years), age-squared, sex (male; female), body mass index (kg/m2) which was categorized into four levels according to Asian cut-off values (underweight: <18.5; normal weight: 18.5–22.9; overweight: 23.0–27.5; obese: ≥27.5) [25], marital status (has a partner; not married; divorced or widowed), educational attainment (illiterate; less than elementary school; junior high school or higher), employment status (not currently employed; farming/fishing; self-employed; formal employee; part-time job with heavy physical activity (e.g., construction workers), part-time job with low-moderate physical activity (e.g., office workers); others), and household income, which was self-reported on a 10-point Likert scale and categorized into tertiles (low; middle; high). Lifestyle factors included alcohol consumption (does not drink; 1 or 2 days a week; 3 to 6 days a week; every day (<50g pure alcohol); every day (≥50g pure alcohol)), smoking (never smoked; stopped smoking; currently smoke), and physical activity. The amount of alcohol was calculated if the participant answered that they consume alcoholic beverages every day. For the usual daily quantity consumed, participants were asked to report types (beer (~4%), rice wine (~25%), strong spirits (~50%), wine (~10%)) and quantity (bottle for beer, which is usually 640 ml in China, and liang (Chinese ounce equivalent with 50g) or for others). Pure alcohol consumption was then calculated and those who consumed 50g or more pure alcohol daily were defined as having heavy drinking habit [26]. Physical activity was measured by a question ‘Compared to other people in your village, how do you rate your own physical activity level?’ on a 10-point Likert scale, and then categorized into three groups (0–3; 3.5–5.5, 6–10) to roughly categorize them into inactive/normal/relatively active groups.

Statistical analysis

After excluding people with missing values on loneliness, hypertension, social support, and covariates (n = 34), the sample size for the analysis was 763. Those who were excluded were more likely to be older, have lower education, have no partner, have lower physical activity, and to be underweight. Infirmity and difficulties in communication were the main reasons for missing data.

A mixed-effect Poisson regression analysis with a random effect and robust variance estimator [27] was used to investigate the association between loneliness, social support and hypertension, since the prevalence of hypertension was high (46%) among the participants [28]. A random effect model was chosen, given the large variations in the levels of urbanization across communities (the proportion of those engaging in farming/fishing: 53.2–80%) and population size (500–1220). Covariates included age, age-squared, sex, body mass index, marital status, educational attainment, employment, household income, alcohol consumption, smoking, and physical activity. Relative excess risk due to interaction (RERI) was also calculated using estimates from the Poisson regression to investigate whether additive interaction is positive or negative [29, 30].

Model 1 analyzed the association between loneliness and hypertension. Models 2 and 3 then included the interaction terms between loneliness and instrumental support and emotional support, respectively. We also conducted sensitivity analyses. First, a least-squares linear regression with a random effect was used to investigate the association between loneliness and hypertension by using log-transformed SBP and DBP as outcomes. Second, covariates were excluded to see the uncontrolled association between main variables. Third, different cut-points were used for loneliness (i.e., frequently and always vs. sometimes, rarely, and never) to see if the findings were robust. Fourth, the age-stratified analysis (younger than 45 years old, 45 to 64 years, and 65 years or older) were conducted. Finally, the analysis with community fixed effects rather than random effects was conducted.

All statistical analyses were conducted using Stata 16.1 (StataCorp, College Station, TX, USA). The level of statistical significance was set at p < 0.05 (two-tailed).

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The protocol was reviewed and approved by the human subjects committees of the Chinese government, Ethics Committee for Medical Research at the University of Tokyo (No. 10515-(1)) and the Ethics Committee of the Institute of Tropical Medicine at Nagasaki University (No. 120910100). Written informed consent was obtained from all individual participants included in the study.

Results

Characteristics of study participants

Table 1 summarizes the characteristics of the study participants. The mean age was 59.0 years, and males comprised 38.9% of the sample. Thirty-nine percent were categorized as lonely, while 46.3% met the criteria for hypertension. The mean SBP and DBP was 140.9 and 82.6 mmHg among those who felt lonely, while it was 133.2 and 79.6 mmHg among those who did not feel lonely. As for social support variables, 70.4% neither received nor provided instrumental support while 24.4% received and provided it, and 48.1% answered they have no friends while 18.4% had 6 or more friends one can turn to for help (i.e., emotional support).

Table 1

Basic characteristics of the study participants (n = 763).

	Total	Loneliness
	(n = 763)	Yes (n = 295)	No (n = 468)
Age (in years)	59.0 [12.9]	61.7 [12.0]	57.4 [13.2]
Sex (Male)	297 (38.9)	106 (35.9)	191 (40.8)
BMI category
Underweight	35 (4.6)	12 (4.1)	23 (4.9)
Normal	350 (45.9)	130 (44.1)	220 (47.0)
Overweight	299 (39.2)	120 (40.7)	179 (38.3)
Obese	79 (10.4)	33 (11.2)	46 (9.8)
Marital status
Has a partner	611 (80.1)	213 (72.2)	398 (85.0)
Not married	29 (3.8)	10 (3.4)	19 (4.1)
Divorced or widowed	123 (16.1)	72 (24.4)	51 (10.9)
Education
Illiterate	285 (37.4)	137 (46.4)	148 (31.6)
Less than elementary school	303 (39.7)	117 (39.7)	186 (39.7)
Junior high school or more	175 (22.9)	41 (13.9)	134 (28.6)
Employment
Not currently employed	189 (24.8)	84 (28.5)	105 (22.4)
Farming/fishing	416 (54.5)	169 (57.3)	247 (52.8)
Self-employment	31 (4.1)	12 (4.1)	19 (4.1)
Formal employee	10 (1.3)	2 (0.7)	8 (1.7)
Part-time job with heavy physical activity	43 (5.6)	9 (3.1)	34 (7.3)
Part-time job with low-moderate physical activity	67 (8.8)	16 (5.4)	51 (10.9)
Others	7 (0.9)	3 (1.0)	4 (0.9)
Household income
Low	335 (43.9)	156 (52.9)	179 (38.3)
Middle	324 (42.5)	99 (33.6)	225 (48.1)
High	104 (13.6)	40 (13.6)	64 (13.7)
Alcohol consumption
Does not drink	564 (73.9)	232 (78.6)	332 (70.9)
1 or 2 days a week	48 (6.3)	20 (6.8)	28 (6.0)
3 to 6 days a week	13 (1.7)	4 (1.4)	9 (1.9)
Every day	71 (9.3)	21 (7.1)	50 (10.7)
Every day (heavy)	67 (8.8)	18 (6.1)	49 (10.5)
Smoking
Never smoked	605 (79.3)	237 (80.3)	368 (78.6)
Has stopped smoking	53 (7.0)	20 (6.8)	33 (7.1)
Currently smoke	105 (13.8)	38 (12.9)	67 (14.3)
Physical activity
Low	241 (31.6)	112 (38.0)	129 (27.6)
Middle	224 (29.4)	79 (26.8)	145 (31.0)
High	298 (39.1)	104 (35.3)	194 (41.5)
Instrumental support
None	537 (70.4)	231 (78.3)	306 (65.4)
Receipt only	27 (3.5)	11 (3.7)	16 (3.4)
Provision only	13 (1.7)	4 (1.4)	9 (1.9)
Both receipt and provision	186 (24.4)	49 (16.6)	137 (29.3)
Emotional support (Number of friends)
None	367 (48.1)	162 (54.9)	205 (43.8)
1 to 5	256 (33.6)	89 (30.2)	167 (35.7)
6 or more	140 (18.4)	44 (14.9)	96 (20.5)

BMI; body mass index. Mean [standard deviation] or n (%) are shown.

Loneliness, social support, and hypertension

Poisson regression analysis revealed that loneliness was positively associated with hypertension (prevalence ratio [PR] 1.12, 95% confidence interval [CI], 0.99–1.26; Model 1 in Table 2). This association was also observed when SBP (coefficient = 0.021, 95% CI, 0.001–0.043) or DBP (coefficient = 0.021, 95% CI, 0.001–0.040) were used as log-transformed continuous outcomes.

Table 2

Results of Poisson regression model with robust variance estimator examining the association between loneliness, social support and hypertension among rural community dwellers in Fujian Province, China (n = 763).

	Model 1	Model 2	Model 3
Loneliness (ref. Low)	1.12 (0.99, 1.26)	1.04 (0.89, 1.22)	1.05 (0.91, 1.20)
Instrumental support (ref. None)
Receipt only		1.08 (0.75, 1.57)	1.07 (0.75, 1.53)
Provision only		1.07 (0.59, 1.95)	1.09 (0.69, 1.73)
Both receipt and provision		0.99 (0.83, 1.19)	1.14 (0.98, 1.33)
Emotional support (ref. Low)
Middle		0.91 (0.80, 1.05)	0.88 (0.75. 1.03)
High		0.83 (0.73, 0.95)	0.72 (0.60, 0.87)
Loneliness x Instrumental support (ref. None)
Receipt only		0.95 (0.58, 1.58)
Provision only		1.08 (0.32, 3.57)
Both receipt and provision		1.49 (1.20, 1.87)
Loneliness x Emotional support (ref. Low)
Middle			1.09 (0.92, 1.28)
High			1.44 (0.90, 2.33)

Values are prevalence ratios and 95% confidence intervals. Covariates included age, age-squared, sex, body mass index category, marital status, educational attainment, employment status, household income, alcohol consumption, smoking and physical activity. A random effects model was used to account for multiple individuals in each community.

The analyses of statistical interaction (Models 2 to 4 in Table 2) showed that relation between loneliness and hypertension was more pronounced among those who reported higher social support compared to those who reported lower support. As shown in Fig 1, in the case of instrumental support, people who did not feel lonely had similar risks of hypertension regardless of the level of support. However, among those who reported feeling lonely, higher instrumental support (both receipt and provision) was associated with an increased prevalence of hypertension. That is, there was a direction of interaction that was opposite to the prediction of the buffering hypothesis. Among people who do not feel lonely, higher levels of emotional support from friends was weakly correlated with lower prevalence of hypertension. However, among individual who felt lonely, everyone converged to a similar risk of hypertension regardless of level of emotional support, i.e., there was no evidence in support of the buffering hypothesis. RERI for the interaction between loneliness and both receiving and providing instrumental support was 0.67 (95% CI 0.39–0.95, p < 0.001) in Model 2, while that for interaction between loneliness and emotional support was 0.17 (95% CI 0.01–0.34) for 1–5 friends, 0.32 (95% CI -0.14–0.78) for 6 or more friends, respectively. That is, the direction of results for additive interaction were consistent with those from the multiplicative interaction.

Fig 1

Interaction of the relationship between loneliness and social support on hypertension.

Models are controlled for age, sex, body mass index, marital status, educational attainment, alcohol consumption, smoking, and physical activity. The y-axis represents predicted probability.

Sensitivity analysis

When we excluded covariates from the analyses, the observed associations were statistically significant (i.e., loneliness was associated with hypertension, and there were significant interactions between loneliness and social supports on hypertension) while the magnitude of the estimates became larger; for example, the association between loneliness and hypertension was stronger (PR = 1.33, 95% CI 1.18–1.51 in Model 1). When we analyzed different cut-points for loneliness (i.e., frequently/always vs. sometimes/rarely/never), the results were attenuated (PR = 1.06, 95% CI 0.89–1.26]) but the overall trends and patterns did not change across models. When we conducted the analyses stratified by age group, the significant association was only observed among those aged 65 years or older (PR = 1.26, 95% CI 1.06, 1.51, n = 269), while there was no significant association among younger people (PR = 0.81, 95%CI 0.44, 1.49 among those aged 18–44 years (n = 105); PR = 1.02, 95% CI 0.81–1.27 among those aged 45–64 years (n = 389)). The interactions between social support and loneliness were also significant among those aged 65 years or older for both instrumental support (loneliness x provision only: PR = 2.05, 95% CI 1.14, 3.71, loneliness x both receipt and provision: PR = 1.47, 95% CI 1.06, 2.06) and emotional support (loneliness x middle support: 1.59 95%CI 1.15, 2.19, loneliness x high support: PR = 1.31, 95% CI 0.99, 1.74). When we ran the analysis with community fixed effects rather than random effects, the effect size for the association between loneliness and hypertension became smaller but trended in the same direction (PR = 1.10, 95% CI 0.96, 1.26).

Discussion

Summary of the findings

In a cross-sectional sample of 763 rural community dwellers in Fujian Province, China, individuals who reported feeling lonely were more likely to have hypertension than those who did not feel lonely, especially among older people. In addition, we found significant interaction between loneliness and social support in relation to hypertension. More specifically, the positive association between loneliness and hypertension was more pronounced among those who reported higher social support compared to those who reported lower support.

Loneliness and hypertension

Our finding in relation to the association between loneliness and hypertension is in line with previous studies. For example, a cross-sectional survey among 1,880 older Malaysians showed that lonely individuals had a higher likelihood of hypertension (6). Also, a longitudinal study among 229 participants in the U.S. found that loneliness predicted increased systolic blood pressure over a 4-year period [5]. One possible pathway linking loneliness and hypertension is the behavioral pathway. Hawkley and Cacioppo [3] have argued that loneliness is equivalent to feeling unsafe and those who feel lonely have higher sensitivity to social threat in the environment; lonely individuals see the social world as a more threatening place, expect more negative social interactions, and remember more negative social information. Social threat can cause diminished self-regulation, so that people with loneliness tend to have worse health behaviors such as more alcohol consumption and less physical activity [31, 32]. In our study sample, those who felt lonely reported lower physical activity (38.0% vs. 27.6%), while physical activity was not clearly associated with hypertension in this study. In addition, less heavy drinking was observed for those reporting loneliness (6.1% vs. 10.5% engaged in daily heavy drinking, defined as 50g or more pure alcohol)). These patterns contradict previous reports in Western settings where lonely people have been found to exhibit unhealthier habits. Given that the inclusion of health behavior variables did not weaken the association between loneliness and hypertension (not shown in Tables), we can at least conclude that in this study, lifestyle differences do not mediate the association between loneliness and hypertension. In rural China, every household typically cultivate their own rice field, and two-thirds of the study participants answered that they engaged in farm work on a daily basis. Hence people in our sample have relatively higher daily physical activity levels than is typical in urban settings, and lower physical activity compared to other people in the same community among the study participants may not necessarily be linked with higher risk for hypertension. In our rural setting, drinking and smoking are also important tools for social interaction, so that the social context of these behaviors also differ from other settings in which loneliness has been studied. A previous study that used data from 29 districts of 3 cities in China found that high membership rate in social organizations was associated with higher prevalence of harmful drinking among both men and women. The authors concluded that the Chinese drinking culture may influence drinking behaviors [33].

Another possible pathway is a pathophysiological pathway. Loneliness can be conceptualized as a chronic stressor. Chronic stress (e.g., long-term activation of the hypothalamic-pituitary-adrenocortical (HPA) axis) has been shown to result in allostatic load [34], resulting in chronic overproduction of stress hormones (e.g., cortisol) which can lead to elevated blood pressure in the long term [5]. In this study sample, loneliness was significantly associated with stress measured by the Kessler Psychological Distress Scale (K6) [35] among those aged 70 years or younger, which was not investigated among older people since there was difficulty understanding Mandarin Chinese (older people in this area usually communicate in dialect) (mean [SD] 3.7 [3.8] vs. 6.7 [5.4]).

Social support and hypertension

As for instrumental support, those with higher social support showed much higher prevalence of hypertension when they were lonely. One possible explanation is that in communities with a comparative absence of working-age adults, those who actively engage in social exchange might also be burdened with a bigger workload and experience greater stress, and thus more likely to establish hypertension. This is so-called ‘dark side of social capital’ [36, 37] which has been previously reported for example, in impoverished communities in the U.S. [38], rural Malawi [39] and rural China [33]. In our previous study which used the same dataset, those who participated more in wedding parties, funerals and social gatherings reported higher stress [22]. Chinese culture (along with other East Asian societies) maintains a tradition of gift exchange for lubricating social relationships with others. Gifts are exchanged between family and friends, but also between co-workers and business associates on significant occasions, such as the Chinese New Year or the mid-autumn festival. In rural areas, gift exchange within social networks functions as a form of informal insurance, creating an obligation on the part of the recipient to reciprocate a favor in the future [40]. For individuals who are feeling lonely, customs surrounding gift exchanges can be a form of social stress [41]. Although we hypothesized that emotional support would buffer the association between loneliness and hypertension, we rather found the opposite trend, viz., among individuals who felt lonely, they converged to a similar risk of hypertension regardless of level of emotional support. Having a lot of friends appeared to be associated with better health (i.e., low risk for hypertension), but this protection was not observed if they felt lonely. We used the number of close friends to whom people could turn to for help when they are in trouble as an indicator of expected emotional support between close friends since a systematic review have shown that a stress-buffering effect is most consistently found when support is measured as a perception that one’s network is ready to provide aid and assistance if needed [42], while it is possible that it also reflects a trait personality factor (e.g., optimism) and bigger social network size. Moreover, as mentioned in the above discussion, the social norms and social context of health behaviors in rural China can be different from those typically found in Western settings. A study in China has shown that those who participate in social activity (e.g., interact with friends, playing mahjong or cards) were less likely to establish hypertension after two years [43]. We found that those who are lonely had low level of social capital (i.e., trust of others, community attachment, and reciprocity among community members) (the unadjusted correlation = –0.11 for trust, –0.17 for attachment, and –0.22 for reciprocity; S1 File), and those reporting higher perceptions of social capital reported more friends (chi-squared test: p < 0.001 for trust, 0.013 for attachment, 0.003 for reciprocity), which may indicate that emotional benefits such as an increase in sense of belonging or purpose through having friends may be more closely linked to hypertension. This may also relate to the finding that there was a main effect of emotional social support on risk of hypertension, which was in accordance with a U.S. study showed that those with emotional support from friends were less likely to have uncontrolled and undiagnosed hypertension [44].

Limitations

This study has several limitations. First, it is not possible to infer causality from cross-sectional design. For example, it is possible that lonely people seek out the emotional support of friends at the same time as developing hypertension (simultaneity), or poor health may lead to increased loneliness and more attempts to seek social support (reverse causality). Previous literature has documented that the amount of received support may reflect poor health status especially in cross-sectional settings [45, 46]. This has been called the support mobilization hypothesis [47], which posits that received support is an indicator of poor health. However, in this study, those who reported feeling lonely were less likely to drink alcohol or to smoke, even though they reported fewer supports. Those who received more instrumental support tended to drink less and smoke, while those reporting more friends tended to be younger, had higher education/income, smoked more, and were more physically active. These patterns tend to suggest that healthier people receive more support (see S1 Appendix). We also note that loneliness could have been the product of social support, i.e., those receiving social support tended to feel less lonely. While we assumed temporal ordering from loneliness → social support, the association between loneliness and social support is likely to be bidirectional. Hence, exchange of social support can be simultaneously a mediator and confounder of the association between loneliness and hypertension. The bidirectionality cannot be teased out in cross-sectional data. It would be also challenging to tease out in longitudinal data, without multiple waves of data capturing changes in loneliness and social support. Second, the participants might not have fully represented adults living in rural communities in Fujian as the survey was conducted in only one city in Fujian and we did not use a random sampling procedure. Especially, it is possible that younger, healthier people left the areas to seek migratory work (healthy migrant hypothesis) and were not included in our survey. Third, the measurement of loneliness relied on one question, although this was because of the relatively low educational background of the participants (37% were illiterate). Fourth, we did not ask receipt and provision of emotional support separately. There are also other aspects of social support to be evaluated (i.e., informational support, appraisal support) and quality of social support was not assessed. Although we tried to address this issue by defining close friendships as those to whom people could turn to for help when they are in trouble, we did not inquire about the perceived quality of support received. Furthermore, our social support measures have not been validated against existing instruments. Fifth, some important variables were not available which enable us to better interpret the findings; for example, sleep quality [3], poor dietary habit especially salt intake and depression, which is distinct from loneliness [48] can be important confounders between loneliness and hypertension. Sixth, the measurement of blood pressure was conducted using an automated oscillometric monitor. Usage of a random-zero sphygmomanometer would have been ideal, but we did not use it due to time and resource restriction. Lastly, the exclusion of those with missing values might have biased the observed associations toward the null, e.g., those with missing values might have had health-related problems which prevented them from engaging in social interactions.

Conclusions

This study showed that loneliness was positively associated with hypertension in rural Fujian communities in China, but the association was not buffered by social support. Overall, feeling lonely in spite of being surrounded by supportive alters in the network was most strongly linked with increased risk for hypertension. Due to the rapid aging of the population and the continued outflow of younger generations to urban centers, it is anticipated that the prevalence of loneliness will rise in rural communities in China [49], and future studies are warranted to address this issue.

Measurement of social capital.

(DOCX)

Click here for additional data file.

Characteristics of social support receiver/provider (n = 763).

(DOCX)

Click here for additional data file.

13 Oct 2021

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Akihiro Nishi, M.D., Dr.P.H.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information.

3. Thank you for stating the following in the Acknowledgments Section of your manuscript:

[AY is financially supported by Japan Society for the Promotion of Science as a JSPS Overseas Research Fellow.]

We note that you have provided funding information that is currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

[This study was financially supported by the JSPS KAKENHI from the Japan Society for the Promotion of Science (JSPS KAKENHI Grant Number JP13J06172; recipient: AY). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.]

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

4. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

5. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

Additional Editor Comments:

Please follow the reviewers comments. I am looking forward to seeing the revision.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: PONE-D-21-28413_reviewer

General comments

Thank you for giving me the opportunity to review this research about the association of the sense of loneliness and blood pressure, which I am sure is an important topic in the aging society. In this cross-sectional study, people with the sense of loneliness were more likely to have high blood pressure, but the statistical association was pronounced once social relationships were adjusted. I have some issues the authors should consider.

Comments

Introduction

1. Line 77–: I understand that the aim of this study was to evaluate 1) the association of the sense of loneliness and high blood pressure in rural China and 2) the role of social relationships in the association. Thus, it seems to me that “the specific mechanisms linking each type of social support and health outcomes remain unclear” is not relevant.

2. Line 82 & 83: I am confused with your usage of “positive” and “negative” in the last sentence of Introduction; I would replace “negative” with “positive.”

Materials and methods

1. Line 89–: You mentioned “the basis of average population size and level of economic development.” I think you should be more specific and explain the criteria you adopted.

2. Line 95–: I am wondering how the difference of the excluded samples and the included ones could impact on the estimates. Could you mention the possible impact in Discussion?

3. Line 116–: I am wondering why only “receiving emotional support” was evaluated and “providing” emotional support was not. Furthermore, how could you claim that the stated number of friends represented just the “reception” of emotional support, instead of reciprocal friendship? I am also concerned about the validity of the three measures of social support.

4. Line 127–: I think income and employment status were important confounders when considering the association of the sense of loneliness and health. Could you include these variables in this study? If you do not have the information, in what direction should the omitted-variable bias be?

5. Line 138–: How about introducing community fixed-effects to account for time-invariant community specific characteristics?

6. Line 148–: I think SBP and DBP were used without making it clear what they stood for. Could you be specific?

Results

1. Table 1: I am wondering why Table 1 was presented according to the age group. I would replace this table by the table in Appendix. Furthermore, what “[]” and “()” indicated should be clear in this table.

2. Line 178–: Why were the mean blood pressure presented here? This information should rather be placed in “Characteristics of study participants.”

3. Line 182: “Fig 1” should be spelled out as “Figure 1.”

4. Line 184: I am not sure whether the expression “lonely people” is appropriate; the sense of loneliness in this study was just a measurement, and some readers might think this expression could induce a form of discrimination.

5. Line 185–: Importantly, we cannot interpret a mere statistical interaction as a biological interaction (a mechanism). How about calculating superadditivity, referring to Chapter 5 in Modern Epidemiology 3rd Edition?

6. Table 2: I would present the estimates and confidence intervals only for the independent variable and interaction terms. You should not present irrelevant rows, all the more so since you did not try to find risk factors.

7. Figure 1: The subtitles of the right two graphs should be different. I would also add confidence intervals for each point.

8. Line 198–: I am wondering what the expression “the observed associations were all unchanged and more obvious” means. Could you be more specific?

Discussion

1. Line 205–: I think you should mention that this study was cross-sectional and the direction of the statistical interaction in the summary paragraph. Furthermore, I would replace the expression “individuals who reported feeling lonely showed a higher prevalence” because prevalence is defined for population, not for a person.

2. Line 215: I do not think the odds ratio of the previous study is necessary unless the figure has a meaning.

3. Line 219: I think you should make it clear what the social threat is.

4. Line 222–: You cannot judge whether people who felt lonely were more likely to be engaged in some healthy behaviors because those associations were subject to confounding. There is a concern for multiple-comparisons, too.

5. Line 222–: The sentence “In our rural setting, drinking and smoking” seems to me subjective. Could you present some evidence to this claim?

6. Line 235–: I am wondering whether you could check the variable for stress; otherwise, it is not clear what this paragraph was aimed for.

7. Line 249–: Could you please explain what the gift-giving culture? Do you have some information on this feature in your sample?

8. Line 250–: This paragraph should be placed on the Limitation subsection. Furthermore, the expression “inverse association” does not make sense; an association can be bidirectional contrary to a causal relationship.

9. Line 261– & Limitation: It is important to keep in mind that this is a cross-sectional study, although a longitudinal study does not solve all the issues; how could you tell social relationships were not a confounder but a mediator? If they were just a confounder, the results were not incompatible with the buffering hypothesis at all. Rather, the results without the adjustment for social relationships might be confounded. How could you justify to claim the association was “pronounced” instead of “less biased” under the possibility of confounding?

Others

1. I have the impression that this manuscript needs a revision by a native English proofreader, although I do not feel qualified to judge English as it is not my mother language. Hence, I might fail to get the gist of this manuscript, especially for Discussion.

2. Conclusion should be concise; new Discussion cannot be included there.

I hope you can make use of the primary data in a clear and transparent way and contribute to the literature on social relationships and health.

Reviewer #2: The authors examined associations of interaction between loneliness and social support with prevalent hypertension and found paradoxical associations: individuals with both loneliness and high social support were more likely to have hypertension than those with either loneliness or low social support. Those findings may be a fact but look like an artifact. I have several concerns below.

Comments

1. According to table 1, participants aged 65 years or older look substantially different from those aged 64 years or younger. Since the late '90s, Chinese economic growth became great, and intranational migration from rural areas to industrial or urban areas drastically increased as the authors mentioned in the introduction. This historical background may cause selection bias and result in paradoxical associations. To avoid such historical selection bias, authors should show the results stratified by age as well.

2. Alcohol intake is assessed only as frequency, not as the amount. Also, the information about quitting alcohol intake was not available. The dose-response association between alcohol intake and incident hypertension has been well-established. However, such an association for prevalent hypertension did not find in the present study. This discrepancy may happen because quitting or reducing alcohol intake was more likely to occur among hypertensive patients and because alcohol intake was misclassified due to the lack of data on the amount of alcohol intake. Furthermore, the present paradoxical findings may be explained by the binge drinking, not assessed in the present study, among individuals with loneliness and high social support.

3. In table 3, women were more likely to have hypertension than men. Is this true? In general, men are more likely to have hypertension than women. The authors need to confirm the reference category of sex in table 3. If no error, there may be selection bias between men and women.

4. In table 3, age is better adjusted for as fifths or other categories than the continuous value because the association between age and hypertension can be non-linear rather than linear. In addition, for BMI, the reference category should have a narrower range such as 18.5-22.9 kg/m2. The overweight criterion for Asians is 23.0 kg/m2.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

6 Jan 2022

We have attached the response letter as a Word file.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

3 Feb 2022

Can social support buffer the association between loneliness and hypertension? A cross-sectional study in rural China

PONE-D-21-28413R1

Dear Dr. Yazawa,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Akihiro Nishi, M.D., Dr.P.H.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

I am happy to accept the manuscript. Please do the final edit (including the one suggestion from the reviewer 2) during the publication process.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: For the last part of Discussion, please consider removing "toward the null" from "Lastly, the exclusion of those with missing values might have biased the observed associations toward the null, [...]"

Reviewer #2: Thank you for carefully addressing my concerns. I do not have any further comments on this manuscript.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Isao Muraki

8 Feb 2022

PONE-D-21-28413R1

Can social support buffer the association between loneliness and hypertension? A cross-sectional study in rural China

Dear Dr. Yazawa:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Akihiro Nishi

Academic Editor

PLOS ONE