A web-based intervention is feasible for supporting weight loss and increased activity in rural women with arthritis.

Rural women have well documented health disparities, with higher prevalence of obesity and chronic conditions, including arthritis. Change in weight and actigraph-recorded data were examined in a subset of 63 of 82 women with physician-diagnosed arthritis who completed a 30-month web-based clinical trial. Repeated measures analyses showed women lost weight from baseline to six months, slowly regained at 18 and 30 months, ending with a lower weight than baseline F(1,62)=40.89, p<0.001, η2p =0.40. Of 53 women with complete data, activity increased at six months, decreased at 18 months, and increased at 30 months F(1,52)=4.14, p =.04, η2p=0.07. Women showed improved change in weight and activity from baseline at six, 18 and 30 months. This study adds support that web-based programs may promote weight loss and activity in a hard-to-reach, underserved population of midlife and older rural women with arthritis.

Introduction

Women living in rural communities are at greater risk of obesity and obesity-related behaviors that contribute a high prevalence of chronic conditions such as arthritis, compared with their urban counterparts.[1,2] Physician-diagnosed arthritis affects nearly one in three adults in rural areas, with approximately half of these adults reporting arthritis-attributable activity limitations.[3,4] The high prevalence of arthritis among rural women is attributed to recognized risk factors of age (45 years or older), lower socioeconomic status, lower levels of activity, and/or higher body mass index (BMI).[3] Attaining modest body weight losses of ≥ 2-5% is related to a reduction in pain and abnormal joint loading,[5-7] and is correlated with health-related quality of life.[8] Maintaining an active lifestyle is important in reducing arthritis-attributable activity limitations and maintaining physical function.[6] Yet, rural women with arthritis are less likely to receive weight loss counseling from their primary care provider than their urban counterparts,[2] potentiating the rural access disparity to health promotion resources.[2]

Purely distance-delivery interventions, via web-based, eHealth, or other formats have shown promise as an effective method to address behavior change for weight loss and weight maintenance in hard-to-reach populations, including rural women, though eHealth studies related to weight management in individuals with arthritis are limited.[9-11] This exploratory study examined whether rural women with obesity and arthritis completing a 30- month web-based intervention would experience change in weight and activity and to assess patterns of change, if found.

Design and methods

This is a secondary analysis of change in weight and activity data over time from a subset of women who self-reported physician- diagnosed arthritis at baseline, and who completed the 30- month Women Weigh-in for Wellness randomized controlled trial (ClinicalTrials.gov Identifier: NCT01307644). All women enrolled in the parent trial provided informed consent, consistent with ethical approval requirements received through the Institutional Review Board of the University of Nebraska Medical Center (approval number 23710-FB). The study was conducted between June 2011 and December 2014. The parent trial randomly assigned 301 eligible women from underserved, rural communities to compare the effectiveness of three interventions: web-based only or web-based supplemented by either peer-led discussion or professional email counseling, on achieving weight loss and weight maintenance. Unique to this trial was the use of a threephased intervention approach to influence initial weight loss and transition to weight maintenance over a 30-month period. The rationale, design and results of this trial have been presented elsewhere. [9]

As the parent-trial data analyses revealed no intervention group differences in the primary outcome of weight,[9] for this study we combined data for analysis from women across all three intervention groups who self-reported physician-diagnosed arthritis at baseline.[12] Women were excluded if they reported a diagnosis of type 1 diabetes or type 2 diabetes requiring insulin, lost 10% or more bodyweight in the past 6 months, participated in another weight loss program, or if they were on medications that affected weight loss or weight gain as verified by research personnel.

All women received the same messaging on the basic intervention website about behavior change strategies for adopting lifestyle healthy eating and activity, following accepted national dietary and activity recommendations at the time of intervention.[13-15] Regardless of group assignment, all women had access to the basic intervention website, whereas women in two groups received augmented features of either participation in a peer-led discussion group or professional email counseling. All women were encouraged to self-monitor using the online web logging feature for weight, daily kilocalorie intake, activity step counts, and goal setting. The intervention was founded on the constructs of Pender’s Health Promotion Model, with delivery of messaging purposely addressing motivational messages about benefits and barriers to action, self-efficacy, and interpersonal influences of family and friends for action of behaviors necessary to achieve weight loss and weight maintenance.[16]

All assessments were completed by trained research nurses at baseline, 6 months, 18 months, and 30 months, at an office on a community college campus, which was centrally located in the geographic region of the women’s residences. Women completed surveys about their general demographic information and health history. Height (m), weight (kg), and body mass index (BMI) were measured via the Tanita scale (TBI-2015, Tanita Corporation of America, Inc.).

An actigraph accelerometer (Model GT3X, Pensacola, FL, USA), a device shown to have high reliability and validity, was used to assess weekly minutes of moderate and/or vigorous physical activity (MVPA), using established cut-points.[17,18] At the assessment time points, women were asked to wear the device placed on an adjustable waist band over their dominant hip. Women were asked to wear the device 24 hours a day, except during activities such as showering, over seven days.

All statistical analyses were completed using SPSS v 25. Repeated measures ANOVAs were conducted to examine women’s change in weight and actigraph-recorded physical activity. Using the four available time points (baseline, 6 months, 18 months, and 30 months), we tested for patterns of change (e.g. linear, quadratic, cubic) over time, if any. Prior to conducting analyses, all variables were assessed and verified to meet adequate standards for normality. Our analysis was limited to 63 women with complete weight data at all four assessment time points.

Results

Sixty-three of 82 women (77% retention) completed the 30- month trial. At baseline, the mean age of the completers was 57.6±6.7 years with an average BMI of 35.0±4.2 kg/m2. Women were non-Hispanic white (n=63, 100%). All women graduated from high school, with 56 (89%) completed some college or greater and 29 reported a baccalaureate degree or higher (46%). Fifty-one women (81%) were employed outside of the home fullor part-time and 52 women (85%) reported an annual household income of $40,000 or greater.

Repeated measures analysis of weight change showed that women lost an average of 5.7±6.4 kg initially between baseline to six months, then slowly regained at 18 and 30 months, and ended the 30 month study with an average of 2.8±6.8 kg lower weight than at baseline, consistent with a quadratic pattern of weight change, F(1, 62) = 40.89, p<0.001, η2p = 0.40 (Table 1 and Figure 1). Mean weight at six months, 18 months, and 30 months was lower than baseline bodyweight. Of the 63 women with complete weight data over time, nearly half lost a minimum of 5% percent of body weight at six months (n=31, 49.2%). Over a third lost 5% or more at 18 months (n=24, 38.1%) and 30 months (n=24, 38.1%). The average mean bodyweight loss from baseline was 5.7±6.4, 3.8±6.8, and 2.7±6.8 kilograms at the respective follow-up assessments. Among 53 women with complete actigraph data, women averaged 213.3±122.3 weekly minutes of MVPA at baseline. At six months, the time associated the greatest weight loss, women substantially increased their activity, averaging 251.7±166.7 weekly minutes of MVPA. Overall the pattern of change in physical activity was cubic as noted in Table 1 and graphed in Figure 2, suggesting an initial increase in activity at six months, decrease at 18 months, and another increase at 30 months, F(1, 52) = 4.14, p=0.04, η[2]p =0.07.

Table 1.

Weight and physical activity of rural women with arthritis (n=63). Analyses using Repeated Measures ANOVAs.

Outcome	Baseline	6 Months	18 Months	30 Months
	M (SD)	M (SD)	M (SD)	M (SD)	F	η2p
Weight (kg)	93.53 (12.96)	87.81 (13.68)	89.72 (14.92)	90.78 (14.59)	40.89**	0.40
Physical activity°	213.28 (122.47)	251.67 (166.68)	217.45 (126.56)	232.84 (146.29)	4.14*	0.07

°Number of weekly minutes of moderate or vigorous intensity activity over 7 days as measured by actigraph. *p<0.05; **p<0.001.

Figure 1.

Average weight (kg) of rural women with arthritis at assessment visits (n=63).

Figure 2.

Average weekly minutes of moderate or vigorous intensity physical activity (MVPA) of rural women with arthritis at assessment visits (n=63).

Women with complete actigraph data were active, exceeding national activity recommendations of ≥ 150 minutes MVPA across all 4 assessment time-points, noting 30 month completers averaged 232.8 ± 146.3 weekly minutes of MVPA, which was higher than baseline activity. Over half of women averaged at least 150 minutes or more of MVPA per week throughout the study (baseline, n=43, 68%; 6 months, n=46, 73%; 18 months; n=41, 65%; and 30 months, n=36, 57%).

Discussion

This analysis adds to the current evidence that a web-based intervention for change in weight and activity was feasible in reaching and retaining an at-risk population of rural women with physician-diagnosed arthritis. Women achieved and maintained a modest weight loss over time, consistent with other weight loss programs using eHealth delivery.[11] Such modest weight losses have resulted in clinical meaningful reduction in joint loads, osteoarthritis symptoms, and improved functional mobility during daily activities in other studies, although these health outcomes were not assessed in this study.[5-7]

The rural women who participated in the actigraph-activity monitoring were active, exceeding national recommendations for ≥ 150 weekly minutes of MVPA across all time points. This level of activity may have influenced their success in maintaining a lower weight than baseline at 18 and 30 months, as activity is associated with weight loss maintenance.[19] This active cohort showed the capacity to substantially increase activity between baseline and six months.

There are limitations in this exploratory analysis. The physician- diagnosed arthritis as self-reported by the women was not verified by a health care provider, yet this method is validated for public health surveillance.[12] Demographics of this cohort, all non- Hispanic white women of higher socio-economic status, may have influenced the results, an issue common in many eHealth interventions. [11,20] Women most likely to be active were potentially more motivated to participate in the actigraph-activity monitoring, and as such may not be reflective of the sample.

Additional research is needed to assess the confounding influences (e.g., comorbidities, pain, psychosocial factors, etc.) affecting women’s change in weight and activity, and potential health outcomes associated with change in these measures. With growing popularity of such distance-delivery interventions, there is a need to involve public health communities in the development of eHealth interventions to advance the acceptability, scalability, and long-term outcomes of the communities they serve.[20]

Conclusions

Findings support the feasibility of reaching and retaining rural women with arthritis to achieve weight reduction and increased activity across 30 months using a web-based intervention. This study broadens the research evidence base with which to inform future public health programming targeting health promotion of rural midlife and older women.