Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
WMJ. Author manuscript; available in PMC 2016 June 1.
Published in final edited form as:
PMCID: PMC4530798

Self-Reported Mental Health Predicts Acute Respiratory Infection

Lizzie Maxwell, BA,a Bruce Barrett, MD, PhD,*,a Joseph Chase, BS,a Roger Brown, PhD,a,b and Tola Ewers, MS, PhDa



Poor mental health conditions, including stress and depression, have been recognized as a risk factor for the development of acute respiratory infection (ARI). Very few studies have considered the role of general mental health in ARI occurrence. The aim of this analysis is to determine if overall mental health, as assessed by the mental component of the Short Form 12 Health Survey (SF-12), predicts incidence, duration or severity of ARI.


Data utilized for this analysis came from the NIH-funded MEPARI and MEPARI-2 randomized controlled trials examining the effects of meditation or exercise on ARI among adults aged >30 years in Madison, Wisconsin. A Kendall tau rank correlation compared SF-12 mental, completed by participants at baseline, with ARI incidence, duration and area-under-the-curve (global) severity, as assessed by the Wisconsin Upper Respiratory Symptom Survey (WURSS-24).


Participants were recruited from Madison, Wisconsin using advertisements in local media. SF-12 mental health scores significantly predicted incidence (p=0.037) of ARI, but not duration (p=0.077) or severity (p=0.073). The PANAS negative emotion measure significantly predicted global severity (p=0.036), but not incidence (p=0.081) or duration (p=0.125). MAAS scores significantly predicted incidence of ARI (p=0.040), but not duration (p=0.053) or severity (p=0.70). The PHQ-9, PSS-10 and PANAS positive measures did not show significant predictive associations with any of the ARI outcomes.


Self-reported overall mental health, as measured by the mental component of the SF-12, predicts ARI incidence.


Acute respiratory infection (ARI), including influenza, comprises one of the most common categories of illness in the United States and worldwide.1 Influenza and non-influenza ARI yield an inordinate economic burden.2 The United States bears estimated annual costs of 40 billion dollars for non-influenza ARI.2 At a time when healthcare is seeking to rein in excessive spending, public health efforts could profit from identifying and targeting factors that increase susceptibility to ARI.

Poor mental health has been implicated as a risk factor for developing ARI.3 Cohen and colleagues identified an association between increased stress and respiratory infection vulnerability. This was demonstrated by an increased likelihood of developing ARI upon viral challenge for people with higher mental or social life challenges.4 A more recent population-based retrospective cross-sectional study revealed that individuals with any diagnosed DSM-IV mental disorder had a 44% greater risk of having developed a cold in the previous 12 months.3 However, such research is in its infancy, and mental health’s impact on ARI susceptibility is still uncertain.3 Despite evidence implicating specific mental conditions such as stress and DSM-IV disorders in the development of ARI, no prospective cohort research has yet looked at whether general mental health influences ARI occurrence.

The Short Form 12 Health Survey (SF-12), a validated instrument measuring generic health-related quality of life, is a reliable measure of overall physical and mental health status.5 Like the SF-36 from which it was derived, the SF-12 can be divided into two summary measures: the Physical Component Summary (PCS-12) and the Mental Component Summary (MCS-12).5 The primary aim of this paper is to determine if general mental health, as assessed by the mental component of the Short Form 12 Health Survey, is correlated with incidence, duration or severity of ARI illness. We will also look at relationships of a variety of other self-report psychosocial measures with ARI illness.



Data utilized for this paper came from two NIH-funded randomized controlled trials: the Meditation or Exercise for Preventing Acute Respiratory Infection (MEPARI) study as well as the first two cohorts of the follow-up MEPARI-2 study. The primary aim of the MEPARI and MEPARI-2 trials was to determine if training in mindfulness meditation or exercise might be effective in decreasing ARI illness burden when compared to the control group.6 The pilot MEPARI trial found positive results, especially for meditation.67 The MEPARI-2 trial is in progress. Detailed methods can be found at (NLM Identifier: NCT01654289). The following serves to briefly describe the methods pertinent to this analysis.

The MEPARI (MEPARI-2) trials enrolled adults ≥ 50 (30–69) years of age. Participants were recruited from Madison, WI by means of advertising in local media. Prospective participants were screened by telephone using a scripted protocol. Following telephone screening, eligible adults were enrolled in a two-week run-in trial to assess ability to adhere to the study protocol. Eligibility criteria included healthy adults who reported having either ≥ 2 colds in the last 12 months or ≥ 1 cold per year on average. Exclusion criteria included moderate exercise ≥ 2 times per week, vigorous exercise ≥ 1 time per week, regular practice or previous training in meditation, autoimmune, immunodeficiency or malignant disorder, a score of > 14 on the Patient Health Questionnaire (PHQ-9) Depression Screen, current or anticipated use of antibiotic or antiviral medications, pregnancy or plans of becoming pregnant, and previous allergic reaction to eggs or the seasonal influenza vaccine.6 Upon successful completion of the run-in trial, participants were eligible for consent and enrollment in the main trial. Participants were randomized to one of three parallel groups: meditation, exercise, or observational control. Those randomized to the meditation and exercise arms underwent eight weeks of training in their respective behavioral intervention. All participants were monitored for ARI occurrence until study exit, with regular contact with study staff, and daily reporting on the Wisconsin Upper Respiratory Symptom Survey (WURSS-24) during ARI illness episodes. Participants in cohort one and cohort two of the MEPARI trial were enrolled for 9 and 5 months, respectively. MEPARI-2 cohorts were enrolled for 9 months each.


Psychosocial Measures

The following psychosocial measures were completed at baseline. The mental component of the SF-12 version 2 (MCS-12) includes five of the twelve SF-12 items, and is calculated using an item-weighted algorithm.8 Three items were derived from the five-item Mental Health Inventory (MHI) assessing common diagnostic symptoms of depression and anxiety disorders, while the other items in the MCS-12 pertain to the level of functional impairment attributable to poor mental health.5 The Positive and Negative Affect Schedule (PANAS) is a valid measure of positive and negative emotion consisting of two 10-item scales.9 PANAS has been used more for research than clinical applications. The Patient Health Questionnaire (PHQ-9) is a self-administered measure used to assess and monitor depression severity.10 Individuals with scores > 14 (moderately severe depression) were excluded from the MEPARI (MEPARI-2) trials. The Perceived Stress Scale (PSS-10) is an instrument assessing psychological stress that has been widely used in clinical and epidemiological research.11 The Mindful Attention Awareness Scale (MAAS) is a well-developed and validated 15-item measure of trait mindfulness.12

Measures of ARI Illness


All participants were monitored for ARI from study onset until study exit, using either biweekly telephone check-in (MEPARI) or weekly electronic surveys (MEPARI-2) and at-home daily self-report during ARI episodes. ARI incidence was determined by study personnel using the Jackson Scale. The Jackson Scale is a sum of symptom scores for the following health outcomes: sneezing, headache, malaise, chilliness, nasal discharge, nasal obstruction, sore throat and cough.13 Each symptom is rated as absent (0), mild (1), moderate (2) or severe (3). An ARI illness episode was defined by a score of ≥ 2 on the Jackson scale, with at least one of the following cold symptoms: nasal discharge, nasal obstruction, sneezing or sore throat. Participants had to answer “yes” to “do you think you have a cold?” or “do you think you are coming down with a cold?”


Total duration of an ARI illness episode was assessed as time from first reported ARI symptom to the last time the person reported being ill. The end of the episode was confirmed by a person marking themselves as “not sick” for two days in a row. Self-report times were recorded in hours and minutes and converted to decimalized days.

Global Severity

The Wisconsin Upper Respiratory Symptom Survey (WURSS-24) is a validated illness-specific questionnaire assessing symptom severity and quality of life impact.14 The WURSS-24 was used to assess daily symptom severity during each ARI illness episode. Items are rated on a 7-point Likert-style severity scale. For each participant, a daily global severity score was calculated by summing WURSS-24 items 2–23.6 Total illness burden was represented by a WURSS-AUC (area-under-curve) global severity score, calculated using trapezoidal approximation across all days of each illness episode (with duration on the X-axis and WURSS-24 severity on the Y-axis).6

Statistical Analysis

Data were analyzed using the R package Version Predictive associations between baseline psychosocial measures and subsequent ARI outcomes were assessed using Kendall’s tau rank correlation. Kendall’s tau rank correlations are unconditional assessments examining the number of concordant and discordant pairs after ordering the values based on ranking each of the quantities.16 Kendall’s tau rank correlation was chosen because of the skewed distributional nature of the dependent variables, duration and global severity, and the binomial nature of incidence.


883 (503) adults were screened for the MEPARI (MEPARI-2) studies; 204 (250) of those screened were entered into the run-in trial, 154 (204) were randomized into the main trial, and 149 (191) were followed through the monitoring period. MEPARI participants underwent assigned 8-week interventions (meditation or exercise) beginning in September 2009 for cohort 1 and January 2010 for cohort 2 (MEPARI-2 interventions began in September 2012 for cohort 1 and September 2013 for cohort 2). Both MEPARI and MEPARI-2 participants were monitored through May of the years in which they participated. A more detailed description of results from the MEPARI trial can be found elsewhere.6 Furthermore, as the MEPARI-2 trial is yet ongoing, the described results pertain only to the first two cohorts of this four-cohort study. Demographic information is presented in Table 1.

Table 1
Demographic information for MEPARI and MEPARI-2 trials

Baseline SF-12 scores in the MEPARI and MEPARI-2 trials ranged from 18.6 to 66.8 with a mean (SD) of 49.8 (8.89). Possible SF-12 mental scores can be as low as 0 and as high as 100, with greater scores representing better mental health-related quality of life. SF-12 mental scores are calculated by item-weighted algorithm, normed to a mean of 50.0 and a standard deviation of 10.0.17 In 2001, mean scores (by age group) in the U.S. were 48.9 (25–34), 48.8 (35–44), 49.9 (45–54), 50.8 (55–64), 51.6 (65–74) and 48.9 (75+).17

ARI outcomes from the MEPARI and MEPARI-2 studies were as follows. There were 282 ARI illness episodes, with a minimum incidence of 0 and a maximum incidence of 5. The mean (SD) incidence per person was 0.81 (0.91) illnesses. Duration ranged from 0 to 85+ days of ARI illness, with a mean (SD) duration of 7.8 (11.3) days. WURSS-AUC (area-under-curve) global severity scores ranged from 0-8724 with a mean (SD) of 298 (656).

The Figure depicts the relationship between SF-12 mental scores at baseline and the number of subsequent ARI episodes. While the sample size for those with 3 or more illness episodes is limited (n=17), the trends are consistent and rather striking.

figure nihms684225f1

Table 2 depicts Kendall tau correlation coefficients for each psychosocial measure with regards to ARI outcomes (incidence, duration and severity). SF-12 mental scores significantly predicted incidence (p=0.037) but not duration (p=0.077) or severity (p=0.073) of ARI. Though not statistically significant, the relationships between duration and severity with SF-12 mental scores trended in the predicted (negative) direction. The PANAS negative measure significantly correlated with global severity (p=0.036) but not incidence (p=0.081) or duration (p=0.125) of ARI. MAAS scores were significantly associated with incidence (p=0.040) but not duration (p=0.053) or severity (p=0.070) of ARI. The PANAS positive, PSS-10 and PHQ-9 measures did not show significant predictive relationships with any of the ARI outcomes.

Table 2
Predictive correlations of mental health scores with ARI outcomes using Kendall tau rank correlations.


This paper provides additional support to the previously reported relationship between poor mental health status and ARI outcomes. Of the six psychosocial measures analyzed, baseline scores for the SF-12 mental, PANAS negative, and MAAS significantly predicted ARI outcomes in the expected directions. Aside from the PHQ-9 and the PANAS positive emotion measure, which showed no discernible trends, the correlation coefficients for the six psychosocial measures with all three ARI outcome measures were in predicted directions. That is, psychosocial measures reflecting better mental health pointed towards fewer, shorter and less severe ARIs, while psychosocial measures reflecting poor mental health were correlated with or trended towards predicting more frequent, longer and more severe ARIs (see Table 2).

The relationship between SF-12 mental health score and ARI incidence depicted in the Figure is especially provocative. Understanding of these associations could be further strengthened with a larger population that included more individuals with three or more colds.

Several potential explanations exist regarding the relationship between poor mental health and increased ARI occurrence and severity. To begin, a common risk factor, such as negative emotion, may increase an individual’s susceptibility to both ARI and poor mental health. Previous studies have demonstrated that in individuals vulnerable to depression, increased negative emotion activates dysfunctional thinking and attitudes,18 which in turn could negatively influence the experience and functional impact of ARI symptoms. Likewise, negative emotion has been shown to interfere with the release of secretory immunoglobulin-A (s-IgA), a primary antibody in the defense against the common cold.18 Therefore, negative emotion may be a causative or mediating agent in the development of both poor mental health and ARI illness.

It is also quite possible that poor mental health may influence ARI outcomes through health-related behaviors. It is documented that rates of smoking19 and excessive alcohol consumption20 are greater among individuals with mental illness. Prior research has shown that such behaviors increase an individual’s susceptibility to respiratory infection by subduing the host immune response.21 Thus, poor mental health may heighten ARI vulnerability by means of health-related behaviors, such as smoking, excessive alcohol consumption or other mediators that impact immunity or susceptibility to infection.

A third consideration is that individuals with poor mental health may merely report more symptoms. In a previous study, for example, it was noted that individuals with a greater negative emotional style reported more unfounded cold symptoms.22 The intensive ARI-monitoring and laboratory verification measures in the MEPARI studies (multiplex PCR viral identification, not shown here) argues against this as a sole explanation for the associations observed, but self-report tendencies and potential biases should not be discounted.

Finally, a complex interplay of the above factors, or as yet unknown causal or mediating pathways, might better explain the observed relations between poor mental health and ARI occurrence.

Our data and analyses have both strengths and limitations. Although past research has investigated the impact of specific mental disorders on ARI outcomes, this paper expands current research by examining non-specific, overall mental health, using validated measures, with a prospective cohort study design. Noteworthy was the use of stringent criteria to define and monitor ARI illness episodes. One of the best previous studies was retrospective, and relied on 12-month recall of ARI incidence.3 Also distinct in our work was the daily assessment of ARI severity; research by Cohen and colleagues assessed cold symptoms for up to nine consecutive days regardless of cold duration,23 whereas subjects enrolled in the MEPARI and MEPARI-2 trials assessed symptom severity daily from illness onset until two days of no reported symptoms, regardless of the actual duration of illness.

An important limitation is the fact that the MEPARI-2 trial is yet ongoing; consequently, this manuscript lacks data pertaining to the final two cohorts of the study, an estimated 200 participants. We expect that incorporation of these final two cohorts will strengthen our conclusions, as the increased sample size will provide greater statistical power. Furthermore, due to the current nature of the study with group status blinded to investigators, we were unable to analyze the three groups (meditation, exercise, and control) separately and therefore our data does not take into account potential confounding effects of meditation and exercise. Having analyzed only baseline mental health indicators, controlling for interventions should not have altered the self-reported mental health scores included in this manuscript, yet the ARI outcomes reported in this paper may have been alleviated by meditation or exercise. As these interventions were assigned randomly and not in relation to mental health, and as any ARI-prevention efforts would reduce the number of illness episodes that mental health indicators could be related to, we do not believe this would invalidate our results.

In addition, the population in this analysis may not be representative of the general adult population. A larger proportion of our study cohort was white (92%), non-hispanic (96%) and college-educated (71.4%) (Table 1) than the general adult population in the United States(2425). Thus it is unclear how these conclusions would translate to other populations.

Finally, despite the significant p-values, the Kendall tau correlation coefficients were small; the extent to which ARI variability is explained by self-reported mental health scores appears to be limited.

In conclusion, using high quality prospective cohort data, we found evidence to support the hypothesis that mental health may influence the occurrence and impact of acute respiratory infection. This is consistent with previous findings, and may have important implications for clinical practice, population health, and public policy. Future studies will be needed to confirm and extend these findings, and to discover ways to reduce impacts of both mental health and respiratory infection.


There was no specific funding for this project. Data came from two trials sponsored by the National Center for Complementary and Alternative Medicine (NCCAM R01AT004313; R01AT006970). Throughout the writing of this paper, Bruce Barrett was supported by a midcareer investigator award (K24AT006543) from NCCAM, which also supports some of Joe Chase's time. Lizzie Maxwell was supported by a Summer Student Research and Clinical Assistantship award from the U.W. Department of Family Medicine, directed by Jonathon Temte MD PhD. The authors are grateful for these sources of support, and would also like to thank the many research participants who provided the data used here.


1. Monto AS. Epidemiology of viral respiratory infections. Am J Med. 2002;112(Suppl-12S) [PubMed]
2. Fendrick AM, Monto AS, Nightengale B, Sarnes M. The economic burden of non-influenza-related viral respiratory tract infection in the United States. Arch Intern Med. 2003;163(4):487–494. [PubMed]
3. Adam Y, Meinlschmid G, Lieb R. Associations between mental disorders and the common cold in adults: A population-based cross-sectional study. J Psychosom Res. 2012;74(2013):69–73. [PubMed]
4. Cohen S, Tyrrell DA, Smith AP. Psychological stress and susceptibility to the common cold. N Engl J Med. 1991;325(9):606–612. [PubMed]
5. Vilagut G, Forero C, Pinto-Meza P, et al. The mental component of the short form 12 health survey (SF-12) as a measure of depressive disorders in the general population: results with three alternative scoring methods. Value Health. 2013;16:564–573. [PubMed]
6. Barrett B, Hayney MS, Muller D, et al. Meditation or exercise for preventing acute respiratory infection: a randomized controlled trial. Ann Fam Med. 2012;10(4):337–346. [PubMed]
7. Obasi CN, Brown R, Ewers T, Barlow S, Gassman M, Zgierska A, Coe CL, Barrett B. Advantage of meditation over exercise in reducing cold and flu illness is related to improved function and quality of life. Influenza Other Respir Viruses. 2012 [PMC free article] [PubMed]
8. Ware JE, Kosinski M, Turner-Bowker DM, Gandek B. User's Manual for the SF-12v2 Health Survey. Boston: QualityMetric; 2008. p. 230.
9. Watson D, Clark L. Development and validation of brief measure of positive and negative affect: the PANAS scales. J Pers Soc Psychol. 1988;54(6):1063–1070. [PubMed]
10. Kroenke K, Spitzer RL, Williams JBW. The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med. 2001;16:606–613. [PMC free article] [PubMed]
11. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. The Journal of Health and Social Behavior. 1983;24:385–396. [PubMed]
12. Brown KW, Ryan RM. The benefits of being present: mindfulness and its role in psychological well-being. J Pers Soc Psychol. 2003;84(4):822–848. [PubMed]
13. Jackson G, Dowling H, Muldoon R. Acute respiratory diseases of viral etiology. VII. Present concepts of the common cold. Am J Public Health Nations Health. 1962;52:940–945. [PubMed]
14. Barrett B, Brown RL, Mundt MP, Thomas GR, Barlow SK, Highstrom AD, Bahrainian M. Validation of a short form Wisconsin Upper Respiratory Symptom Survey (WURSS-21) Health Qual Life Outcomes. 2009;7:76. [PMC free article] [PubMed]
15. R Core Team. R Foundation for Statistical Computing. Austria: Vienna; 2014. R: A language and environment for statistical computing. URL
16. Kendall M. A New Measure of Rank Correlation. Biometrika. 1938;30(1–2):81–89. JSTOR 2332226.
17. Interpreting the SF-12. 2001 Retrieved Aug. 13, 2014, from,
18. Miranda J, Gross J, Persons J, et al. Mood matters: negative mood induction activates dysfunctional attitudes in women vulnerable to depression. Cognit Ther Res. 1998;22(4):363–376.
19. Lasser K, Boyd W, Woolhandler S, et al. Smoking and mental illness: a population-based study. JAMA. 2000;284(20) [PubMed]
20. Kuo P, Gardner C, Kendler K, et al. The temporal relationship of the onsets of alcohol dependence and major depression: using a genetically informative study design. Psychol Med. 2006;36:1153–1162. [PubMed]
21. Cohen S, Tyrrell D, Russell M, et al. Smoking, alcohol consumption, and susceptibility to the common cold. Am J Public Health. 1993;83(9):1277–1283. [PubMed]
22. Cohen S, Doyle W, Turner R, et al. Emotional style and susceptibility to the common cold. Psychosom Med. 2003;65(4):652–657. [PubMed]
23. Cohen S, Miller G. Stress, immunity, and susceptibility to upper respiratory infection. In: Ader R, Felten D, Cohen N, editors. Psychoneuroimmunology. 3rd Ed. Vol. 2. New York: Academic Press; 2001. pp. 499–509.
24. U.S. Census Bureau. Educational Attainment in the United States: 2003. 2004 Retrieved from:
25. U.S. Census Bureau. Resident Population by Sex, Race, and Hispanic Origin Status: 2000–2004. Statistical Abstract of the United States: 2006. Retrieved from: