Poor relationships damage our health.  Recent research powerfully demonstrates this point (Stinson, Logel et al. 2008).  In these studies, relationships were assessed in three different ways - relationship quality (closeness, trust, satisfaction), number of friends, and relationship stress.  Sheldon Cohen (Cohen 2004) has argued that these three aspects of relationships are all important in the relationships-health link - emotional closeness, broader social network, and low interpersonal conflict.  In this Stinson et al research, all three aspects were assessed and all three predicted subsequent health.  In the team's second study, they showed relationship stress (function) and number of friends (structure) were independently linked to health outcomes - the former a bit more strongly than the latter.  More stress and fewer friends both predicted more health difficulties.  Health difficulties too were assessed in three different ways - simply by asking participants whether they had developed any health problems during the study period, by asking about time off work, and by asking about visits to doctors.  Poor relationships led to increases in all three of these health indicators.

Study one followed 135 first year college students over about 3 months.  As they'd expected, the researchers found that lower initial self-esteem led to deterioration in reported relational quality over the 3 months of the study (see questionnaires below) which, in turn, led to more reported health problems.  Study two tracked 188 first year students over their initial 10 week semester using fortnightly online questionnaires.  Lower starting self-esteem led to more relationship stresses and fewer reported friends which, in turn, led to more missed lectures and more visits to doctors.  The vicious circle then tended to continue with poorer relationships feeding back to reduce subsequent self-esteem even further, and so on. 

In this second study, the researchers found that LSE's (people with low self-esteem, defined here as scoring a standard deviation below the mean level for self-esteem - less than 50 in total on the Rosenberg scale below) missed 1.16 more classes per month and visited the doctor 0.39 times per month more than HSE's (people with high self-esteem, defined here as scoring a standard deviation above the mean level for self-esteem - more than 80 on the Rosenberg).  By taking average costs for college classes and medical visits, one can generalize these results.  A conservative estimate is that lower self-esteem and poorer relationships in first year college students in the United States costs $127 million per month for missed classes and extra visits to the doctor.  It's fairly mind-boggling to think of what figures might emerge if research generalized these estimates for work loss and medical costs to the general population.

In a more practical way, what can one do if one finds one has low self-esteem?  There are many options here.  One is to realize that self-esteem isn't all good by any means - see Bauermeister et al's critique.  A second is to let others' appreciation in more.  Yet another is to focus out more than focusing in - see the fascinating work by Crocker et al.  For further ideas, do a tag search on this website using words like self-esteem, compassion, and wellbeing.

Cohen, S. (2004). "Social relationships and health." Am Psychol 59(8): 676-84.  [PubMed]
Relationship Questionnaire - Stinson et al assessed social support, social integration, and interpersonal conflict.  There are reasons to predict that these are three particularly important aspects of relationships, and that they may each make independent contributions to the relationships-health link.  To download a copy of this three aspect Relationship Questionnaire, click here.  Note the second page gives scoring information.
Rosenberg Self-Esteem Scale - in the Stinson et al paper self-esteem was initially rated using just 4 questions from Rosenberg each marked on a 7 point scale.  In the second study, the full 10 question scale was used marked on a 9 point scale (from 1, strongly disagree, to 9, strongly agree).  To download a copy of this fuller questionnaire, click here 
Stinson, D. A., C. Logel, et al. (2008). "The cost of lower self-esteem: testing a self- and social-bonds model of health." J Pers Soc Psychol 94(3): 412-28.  [PubMed]