Family Relationships and Diabetes Care During the Adult Years

Clinical lore, backed by extensive literature, has supported an ongoing interest among clinicians in how family relationships affect and are affected by diabetes. Nowhere has this interest been more attended to than during childhood and adolescence, and a growing interest is evolving in caring for patients who are elderly. There is, however, a notable gap in clinical attention concerning how family relationships influence diabetes management during the adult years.1

This is somewhat surprising because the mean age of diagnosis of type 2 diabetes falls in the fifth decade of life, and a growing number of new patients are being diagnosed in their 20s, 30s, and 40s.2 In this article, I argue that family and couple relationships during the adult years are as important for diabetes management and outcomes as they are for individuals with diabetes who are children, adolescents, or elderly. Furthermore, I suggest that educational and behavioral intervention programs need to address the family context of disease management among adults as they do for patients in other developmental life stages.

I pose three questions to guide the discussion. First, what aspects of family and couple relationships are linked to the management of diabetes during the adult years? This is an important question for two reasons: documenting links between the family and diabetes also identifies important risk and protective factors that can define high-risk family settings for early intervention. Knowing what aspects of family life are tied to diabetes outcomes also identifies targets for intervention. For example, if we find that poor conflict resolution skills between marital partners around diabetes issues are linked to poor clinical outcomes, poor conflict resolution becomes both an indicator of early risk and an area in which we might want to intervene with a structured conflict resolution skill-building program.

Second, what are the documented health risks for family members of people with diabetes? If diabetes is truly a family affair, what are the health consequences for non–genetically related partners of having diabetes in the family?

Third, if both the first and second questions yield convincing information, what do we know about developing and implementing interventions to address the context of the patient-partner diabetes-related relationship?

It might be best to begin by defining “family” for this age period. We will assume that family includes people with diabetes and their partners between the ages of 18 and 25 years at the low end and 60 and 65 years at the high end. Family, for our purposes, includes two cohabiting adults, legally married or not, and of the same or different sexes. Beyond this brief definition, family is what people with diabetes and their partners say it is, which acknowledges the broad diversity of family configuration and type.

What Family Factors Are Linked to Diabetes Management and Outcomes?

Several studies have addressed this question, and a list of documented family risk indicators for this age period is presented in Table 1. The items with an asterisk display the strongest and most consistent links with diabetes management and outcomes across studies. Most of the studies are cross-sectional, but a few have been longitudinal.3,4 They show that couples who score highly on these indicators display progressively poorer diabetes management and outcome over time. The risk indicators are consistent across both type 1 and type 2 diabetes, patient sex, and, for the most part, ethnicity.

Table 1.

Couple/Family Risk Factors During the Adult Years

The most powerful couple risk indicators focus on nonconcordant diabetes beliefs, negative emotional tone, poor problem-solving and conflict resolution skills, and low organizational abilities. The identified family risk indicators for diabetes in this age range are areas where focused relationship-based interventions might be directed.

What Are The Health Risks to Partners of People With Diabetes?

I have divided risk to partners into three areas: biological risk, health behavioral risk, and affective/emotional risk. It is interesting to note that there are relatively few studies in the adult age range that consider the health risk of diabetes on other family members, although there are a considerable number of such studies for family members of child, adolescent, and elderly patients.

Biological risk

The correlation between people with diabetes and their partners on BMI generally ranges between 0.34 and 0.38, and partners of patients with diabetes have a significantly higher BMI and triglyceride level than partners of people without diabetes.5 Furthermore, the same study5 showed that non–genetically related partners of people with diabetes have a 2.11 increased risk of being diagnosed with diabetes themselves during their life-time than control partners without diabetes in the family. Likewise, one in five partners of people with diabetes display evidence of glucose intolerance.5 These findings suggest substantial biological risk to partners of people with diabetes.

Health behavioral risk

This area addresses the relationship between patients and partners on lifestyle behaviors, such as diet and physical activity. It stands to reason that people who live together for the most part share at least some behavioral lifestyle practices, which may present increased risk to partners.

Based on a multiethnic sample of > 500 patients with type 2 diabetes and their partners from our laboratory,68 we found the following significant correlations between patient and partner scores: consumption of fat as a percentage of total calories, r = 0.42; consumption of carbohydrates as a percentage of total calories, r = 0.27; total physical activity level over an average week, r = 0.16; and smoking and alcohol usage, r = 0.22. These correlations fall into the moderate range, but considering the biological risk mentioned above, we view these correlations as clinically significant indicators of health behavioral risk.

Affective/emotional risk

How does having diabetes in the family affect partners' mood, quality of life, and level of depressive affect? Given the important effects of diabetes on these outcomes for patients, we expect that partners might react to the ongoing stresses and strains of chronic disease management in the home in a similar way.9,10 Interestingly, there are very few studies of this issue in the literature for diabetes in the adult years.

Using data from our laboratory on the same sample of patients with type 2 diabetes and their partners mentioned above, we found the following correlations between patients and partners on the Diabetes Quality of Life subscales: satisfaction, r = 0.38; impact on life, r = 0.48; and worry about diabetes, r = 0.25. Furthermore, using the Center for Epidemiological Studies–Depression scale, a 20-item questionnaire that assesses depressive affect, we found that the mean level of partner depressive affect was quite similar to the elevated levels of patient depressive affect.1113

Interestingly, for female partners, the mean level of depressive affect was as high as or even higher than that of patients; for male partners, however, the mean level was lower than that of patients but still higher than that of control subjects. Also, the same factors that were linked to high depressive affect in patients, such as patient functional limitations, were linked to high depressive affect among partners.

Considering all three areas of risk (biological, health behavioral, affective/emotional), the conditions associated with having diabetes in the family pose substantial health risks to non–genetically related partners as well as to patients. This is a rarely documented health consequence of diabetes.

What Do We Know About Couple/Family Interventions in Diabetes?

Given the course of this discussion so far, it should come as no surprise that although many intervention studies in the clinical literature include partners of adult patients with chronic conditions and of families with diabetes among children, adolescents, and the elderly, there are almost no studies of partners and patients with diabetes in the adult years.14 A review of the existing literature15 suggests that studies of adults with diabetes that address the following two crucial questions are sorely needed: What is the effect of including partners of adult patients with diabetes in educational/behavioral programs? What is the effect of addressing relationship issues versus addressing only traditional disease knowledge and management issues in such programs? Ideally, we might want to use a simple 2 × 2 design as outlined in Figure 1 to address these two related questions in a systematic way.

A review of the literature during the past 10 years indicated that no single study explicitly used this kind of design, but a few approached it with very interesting results. In a now-famous study of a weight-loss intervention by Wing et al.,16 the effect of partner attendance was assessed, with partners included in one arm and not in the other arm of the study. They found an important sex-specific effect: women did better if their partners attended, whereas men did better if they attended alone. In a more recent study of a similar weight-loss intervention, Gorin et al.17 showed that including partners in the intervention made a positive difference regardless of patient sex, but only if the partner was also successfully and actively engaged in the weight-loss program.

These two studies suggest the importance of collaboration in a shared couple endeavor, both actively working together to reach a common goal. Interestingly, these studies addressed only one of the two questions posed by the 2 × 2 design: the inclusion versus noninclusion of partners. The inclusion of relationship issues was not addressed in either of the two studies.

Three other studies have been completed with adult patients with diabetes and their partners that come close to the two design questions raised above. All three are unpublished pilot studies, and I express appreciation to the investigators for their permission to share some of their findings. Schermer developed a 6-week traditional diabetes education program with two arms: one included patients and partners and the other only patients (J. Schermer, unpublished observations). No relationship issues were addressed. Trief used two active intervention arms and a control arm. Couples attended both interventional arms: in one the relationship-focused intervention was delivered face-to-face, whereas in the other the intervention was delivered via telephone (P. Trief, unpublished observations). Polonsky developed a two-arm trial, and partners attended both arms. Relationship issues were included in one arm but excluded in the other (W. Polonsky, unpublished observations). In combination, these three studies address one or the other of the two design questions posed above.

Figure 1.

Proposed design for studies addressing issues related to including partners in and addressing relationship issues as part of diabetes interventions.

These studies were pilot projects with small samples. It is noteworthy that in all cases the patients and partners in the active intervention arms did better than control subjects, although many of the results were not statistically significant. These data are only suggestive, but they point to a line of clinical intervention that is of considerable importance for the large number of people with diabetes and partners in the adult years.


To place the findings reviewed in this brief report in context, it is helpful to mention a very interesting caveat from the pilot study by Polonsky mentioned above. He intervened with a 1-day relationship-focused couples workshop and did a 3-month follow-up. He found that those couples who were doing well with each other regarding diabetes at the outset displayed significant improvements at follow-up relative to control subjects. However, couples who were just doing OK or who were not doing well at all at the outset showed either no gains or showed some negative results at follow-up.

This problematic finding raises an important clinical issue, one that Peyrot et al.18 have called the “double-edged sword” of working with adult couples. On the one hand, there are clearly understandable benefits to helping couples work together toward better diabetes management. These include increased collaboration and support, more shared beliefs and activities, better joint problem solving and conflict resolution, and reduced interpersonal distress. On the other hand, addressing relationship issues in a diabetes intervention program has been complicated historically. Partners have been, for the most part, passively not included in diabetes educational and behavioral programs by some practitioners and, interestingly, by some patients.

The focus of almost all diabetes intervention programs for adults has been on individual patients, and for good reason. There are longstanding, powerful, and often unacknowledged feelings among both patients and partners regarding differences in diabetes-related beliefs, issues of control, worries and concerns, needs for autonomy, and fears of the “diabetes police.” These feelings and attitudes are grist for the interventional mill, but their longstanding unacknowledged status sometimes causes them to play out in powerful ways once the door is opened and the elephant in the room becomes visible. Clinicians implementing new family-oriented programs for adults with diabetes and their partners will have to pay close attention to these longstanding, neglected relationship issues that often underlie how couples manage diabetes.

This process, although crucial, can complicate the intervention and cause hesitancies among educators and clinicians who fear stirring up a hornet's nest. Yet, we have already documented the negative impact of couple/family risk factors on diabetes management and outcomes, and we already know that partners are at documented high risk for serious health consequences of having diabetes in the family. We have the technology to engineer systematic educational and behavioral programs to address couple and family issues for people with diabetes and their partners in the adult age range, but we do not as yet have an established history in behavioral diabetes research to address these issues pragmatically.

Addressing this gap will require a shift in perspective and a willingness to bring into the intervention what many practitioners and patients have passively allowed to remain unaddressed. The need is apparent and the focus has been neglected far too long.


  • Lawrence Fisher, PhD, is a professor in the Department of Family & Community Medicine at the University of California, San Francisco.


| Table of Contents