Lifestyle Medicine Assessment Scores in Family Medicine Providers

Abstract

Background: A 21-question Lifestyle Medicine Assessment (LMA) tool was recently developed to quickly evaluate the lifestyle medicine domains (Avoidance of Substance Use, Nutrition, Connection, Movement, and Recovery). In this study, we used the LMA tool to complete a preliminary analysis of family medicine providers, a group that is known to face challenging lifestyle demands and burnout. Objective: The primary purpose was to assess the LMA domain scores and determine if significant differences exist (among the domains). The secondary purpose was to complete a correlation matrix for the LMA domain scores to better understand the strength of the relationships (among domains). Methods: The participants were 35 family medicine providers and all completed the LMA. The maximum LMA score is 50 (each of the five domains is scored 0-10 points). Pairwise comparisons were used to determine if there were significant differences among the five domain scores (P < .05) and Pearson correlation was used to assess correlations between domains. Results: The overall LMA score was moderate (34.68 ± 6.26). The Recovery domain score (4.71 ± 2.91) was significantly lower (P < 0.01) than the other domains (Connection = 6.94 ± 2.16; Nutrition = 7.41 ± 2.00; Movement = 6.26 ± 2.94; Avoidance of Substance Use = 9.35 ± 1.28). The Avoidance of Substance Use score was significantly higher than the other domains (P < .01). Overall, the correlation scores among the domains were weak (all r-scores were ≤0.34). Conclusions: For this group of providers, there appears to be substantial room for improvement in four of the five domains. The domain of greatest concern appears to be Recovery. The weak correlation scores suggest that domain scores tend to be independent of each other, and that assessment results are unique to the individual.

“Most of our providers reported managing stressors effectively most days, which may also help explain why the substance use rates were low.”

Introduction

A 21-question lifestyle medicine assessment (LMA) was recently developed to quickly assess the six pillars of health defined by the American College of Lifestyle Medicine: nutrition, physical activity, restorative sleep, social connection, avoidance of risky substances, and stress management. ¹ The LMA was first fine-tuned with feedback from experts and patients, followed by a formal validity assessment. Strong results were reported for the item content validity index (≥.91), face-validity index (≥.81), scale-level content validity index/face-validity index for relevance (.99), and clarity (.95). The LMA is important to thoroughly study as the American Academy of Family Physicians recently encouraged its’ use (through member resources) and through the American College of Lifestyle Medicine. Further, the field of lifestyle medicine has become increasingly popular as a cost-effective modality to successfully treat, prevent, and reduce the burden of chronic disease.^2-4

For the LMA, the maximum score is 50 points, with each domain having a maximum score of 10 points (connection, movement, substance use, recovery, and nutrition). This questionnaire was purposely designed to assess all domains at once; previously, each unique questionnaires were used for each domain.^5-10 The LMA provides new opportunities to study many groups in need including military personnel, physicians, trainees, nursing staff, the general population, and clinical populations.^11-14 In this paper, we focus on family medicine providers, a group known to a group that is known to face challenging lifestyle demands and burnout.

While lifestyle assessments are routinely recommended for patients, healthcare providers themselves are a particularly vulnerable population often requiring demanding work schedules and high stress. ¹² Each lifestyle medicine domain has been studied individually in physicians and other medical providers, highlighting the need for research to encompass all areas.

Stress Management and Restorative Sleep (Recovery)

It is well-understood that burnout is common in physicians. Approximately 21.7% of healthcare professionals experience high levels of burnout, especially those sleeping 7 hours or less, regardless of gender. ¹⁵ This burnout manifests as an increased risk of mental health disorders, including depression and anxiety, while physical symptoms such as chronic fatigue and muscle pain can further compound the problem.^13,15,16 The consequences extend beyond personal health, as stressed medical professionals are more prone to making mistakes and experiencing work-related accidents, potentially compromising patient care. ¹⁷ Healthcare providers who maintain healthy lifestyles are not only less susceptible to burnout and medical conditions, but are also more likely to effectively counsel patients on health improvement strategies. ¹⁸ This dual impact on both provider wellbeing and patient care quality makes physicians a particularly crucial group for lifestyle intervention research.

Further, healthcare providers often suffer from sleep deprivation. Providers typically average 6.5 hours of sleep per night which is short of the recommendation of at least 7 hours of sleep to maintain overall good health.^19-21 This can become problematic as lack of sleep can contribute to burnout but also lead to medical errors in patient care. ²² Saintilia et al. ¹² showed that healthcare professionals from a public sector hospital in Peru who slept < 7 hours during workdays and days off were 8.33 times more likely for men and 17.18 times more likely for women to have burnout compared to those who reported ≥7 hours of sleep.

Physical Activity (Movement)

The weekly exercise recommendations, according to the American College of Sports Medicine (ACSM) and American Heart Associate (AHA) are: (1) at least 150 minutes of moderate-intensity aerobic activity and (2) at least 2 sessions of moderate to high intensity resistance training. ²³ Those who engage in physical activity beyond 300 minutes (5 hours) weekly have been shown to have substantial health benefits ²⁴ as there is a dose-response association between sitting time and all-cause mortality. A review of the 2020 National Health Interview Survey (NHIS) showed that only 24.2% of US adults met the 2018 Physical Activity Guidelines for Americans for both aerobic and muscle-strengthening activities.^25,26 A study looking at the US Department of Health and Human Services (DHHS) 2008 guidelines have shown that physicians and medical students are more active compared to the general US population ²⁷ ; 64.5% of the general US adult population were active vs 78% of medical student or physician survey respondents. ²⁵

Nutrition

Overall, there is limited nutrition education in postgraduate medical training. ²⁸ Consequently, providers tend to have limited knowledge of nutrition guidelines. For example, Aggarwal et al. ²⁸ revealed only 25% of physicians (n = 300) knew the AHA recommendations for fruits and vegetables. A study on nurses and physicians in Poland concluded only 12.2% of the study participants engaged in healthy eating habits. ²⁹ Physicians have reported a lack of time for adequate nutrition and self-care.^28,30 Further, physicians have report inadequate nutrition available in the workplace. ³¹ For nurses, unhealthy eating habits worsened with longer years worked. ²⁹ These factors can result in both deficits for personal nutrition habits in medical professionals as well as nutrition counseling for patients. ²⁸ Nutrition appears to be an area needing improvement.

Avoidance of Substance Use

The use of risky substances can vary among medical specialties.^32,33 The 2015 study of Oreskovich showed 15.3% of the 7209 physician respondents had met criteria for alcohol abuse or dependence which was similar to rates seen in the general population.^33,34 A meta-analysis from Besson et al. ³⁵ showed that the prevalence of smoking among physicians was around 21% with higher rates seen in family practitioners and medical students. Contributing factors studied in healthcare workers include work stressors, job dissatisfaction, burnout symptoms, and sleep difficulties.^33,36 Alcohol use in physicians was associated with burnout, depression, suicidal ideation, lower quality of life, lower career satisfaction and medical errors. ³³ Providers appear to be susceptible to risky substances like the general population.

Connection

Social connection is an integral part of physical, mental, and spiritual wellbeing. Social isolation can contribute to a 50% increased risk for developing dementia, 400% increased risk of death in heart failure patients, 39% increased of CAD, and 32% increased risk of stroke. ³⁷ Loneliness is when an individual lacks meaningful interpersonal relationships or interactions ³⁸ A recent study reported a loneliness prevalence of 43% for physicians. Prevalence of loneliness is associated with burnout and depression. ³⁸ This is a growing concern amongst healthcare providers, and more attention is needed to study and address this issue.

Objectives

The primary purpose was to assess the LMA domain scores in family providers and determine if significant differences exist (among the domains). The intent is to better understand if the providers tend to excel or struggle with each domain. We hypothesize there will be significant differences among the domains, with recovery having the lowest domain score. The secondary purpose was to complete a correlation matrix for the LMA domain scores to better understand the strength of the relationships (among domains). Strong correlations would suggest individuals tend to perform similarly across domains while weak scores would suggest there are no relationships across domains. We hypothesize there will be moderate to strong correlation scores among the domains.

Methods

The participants were 35 Mayo Clinic providers including family medicine physicians (n = 24; males = 9, females = 15) and Advanced Practice Providers (APPs) (n = 11, female = 11). The consent form and LMA questionnaire were completed electronically (via email) from late March to early April of 2024.

The LMA consolidates the six lifestyle medicine pillars into five domains: Connection, Movement, Nutrition, Recovery, and Avoidance of Substance Use. Note that in the LMA, the lifestyle medicine pillar of restorative sleep is combined with stress management to create the recovery domain. Each domain has multiple questions, with scores weighted based on their impact on morbidity and mortality. The maximum score for each domain is 10 points, resulting in a total maximum score of 50 points. Higher scores indicate healthier lifestyle behaviors. The total score is then interpreted using the following four performance levels: Below Average (0-20), Average (21-30), Very Good (31-40), and Excellent (41-50). For the purpose of this project, we used the same scoring scale for the 10-point individual domain scores: Below Average (0-4.0), Average (4.01-6.0), Very Good (6.01-8.0), and Excellent (8.01-10.0).

For the primary purpose, pairwise comparisons were used to determine which lifestyle medicine domain means were significantly different from each other. For the secondary purpose, Pearson correlation was used test to determine the strength of correlations (r-score) among the five domains.

Results

To present the results and facilitate a better understanding of the scores, we developed a color-coded system illustrated in Table 1.

Table 1.

Color-Coded System Created to Represent Individual Domain Scores and Overall Scores.

Color System Coded for Scores
Each domain individual results		Overall score results
0 - 4.99	Below average	0-20	Below average
5.0 - 6.99	Average	21-30	Average
7.0 - 8.99	Very good	31-40	Very good
9.0 - 10	Excellent	41-50	Excellent

Overall Results

Table 1 presents the numeric and color-coded systems used to present domain scores and total scores.

Table 2 displays the overall results for the 35 providers. The overall LMA score was 34.57 ± 6.20. Therefore, the average deduction was 15.43 points. According to the LMA scoring, this is near the middle of “very good,” which is defined as 31-40 points. Overall scores were highly variable for the 35 providers, ranging from 21 to 46. Nine providers had an “average” score (21-30), 17 had “very good” scores (31-40), and 8 had “excellent” scores (41-50). No providers had “below average” scores (0-20).

Table 2.

Individual Results.

Subject #	Connection	Movement	Substance Use	Recovery	Nutrition	Total
1	6	2	10	0	4	22
2	8	7	6	5	5	31
3	10	8	10	5	8	41
4	6	7	8	5	5	31
5	8	9	10	4	4	35
6	6	7	10	5	8	36
7	6	3	10	0	8	27
8	8	9	10	7	8	42
9	10	10	10	6	6	42
10	4	5	6	0	6	21
11	8	3	10	10	6	37
12	8	10	10	8	10	46
13	8	0	10	5	6	29
14	8	3	10	7	8	36
15	4	10	8	8	9	39
16	6	5	10	2	9	32
17	8	8	10	10	5	41
18	10	8	10	8	7	43
19	4	4	10	4	8	30
20	6	8	8	0	10	32
21	8	10	6	3	10	37
22	8	4	10	5	4	31
23	10	7	8	8	10	43
24	6	8	10	5	9	38
25	4	5	10	4	10	33
26	6	7	10	2	9	34
27	6	0	10	2	8	26
28	4	5	6	8	7	30
29	8	2	10	2	7	29
30	8	6	10	9	10	43
31	4	8	10	4	6	32
32	2	5	10	2	9	28
33	10	7	8	5	8	38
34	8	10	10	5	7	40
35	10	10	10	2	3	35
Mean	6.97	6.29	9.26	4.71	7.34	34.57
SD	2.13	2.90	1.38	2.87	2.01	6.20

Results for Primary Purpose

Figure 1 displays the results for the comparison of domain scores (primary purpose). As hypothesized, there were significant differences among the domain scores. The providers scored the worst on the Recovery domain (4.71 ± 2.87); this was significantly lower than the other four domains (P < 0.01) and it was the only domain score in the “average” category (4.01-6.0). Three of the domain scores were in the “very good” category (6.01-8.0); this included Movement (6.29 ± 2.90), Connection (6.97 ± 2.13) and Nutrition (7.34 ± 2.01). There were no significant differences among these three domain scores. The providers scored the best on the Avoidance of Substance Use domain (9.26 ± 1.38); this was significantly greater than the other four domain scores (P < 0.01) and it was the only score in the “excellent” category.

Figure 1.

Lifestyle score results by domain.

Results for Secondary Purpose

Table 3 presents the correlation matrix for the five domains. We expected moderate to strong correlations among the domains. This hypothesis was rejected as the correlations scores were all weak (all r-scores were ≤0.35). Figure 2 displays an example correlation plot (recovery and movement).

Table 3.

Correlation Matrix for the Five Domains.

	Recovery	Movement	Connection	Nutrition	Substance Use
Recovery	x	r = 0.26	r = 0.35	r = 0.06	r = 0.05
Movement	x	x	r = 0.07	r = 0.09	r = 0.17
Connection	x	x	x	r = 0.20	r = 0.13
Nutrition	x	x	x	x	r = 0.05
Substance use	x	x	x	x	x

Figure 2.

Example correlation plot between domains (recovery vs movement).

Figure 3 displays the distribution of provider scores for each domain. For Avoidance of Substance Use, the majority of providers scored Excellent (26/35) or Very Good (5/35) and only four had Average scores. For the other domains, Below Average and Average scores were far more prevalent (14+ providers).

Figure 3.

Distribution of provider scores for each domain (N = 35).

Results for Single Questions

Table 4 displays the questions for the LMA that the providers performed best and worst on. The providers had excellent scores for 7 of 21 questions, meaning the majority received the maximum awarded points (for that question). Conversely, the providers had poor scores on 3 of 21 questions, meaning the majority received no points (for that question). The scoring criteria for these questions can be reviewed on page 2 of Appendix 1.

Discussion

Group Results

Regarding group findings, the LMA tool is useful as it can help to quickly identify domains of concern for the group (i.e., the majority of the cohort had a low score). For the providers in this study, there was is a clear concern: the Recovery domain. This was the only domain with an Average score (4.71/10). In fact, the majority of providers (25/35) had Average or Below Average scores in this category; this appears to be the one domain where a group intervention may be most needed/appropriate. Three of the remaining domains (Movement, Connection, Nutrition) had means that were Very Good (6.29-7.34). These scores were higher than we expected. For these three domains, at least half the providers had Very Good or Excellent scores. Therefore, group interventions may not be needed/appropriate; it may be better to focus on providing support at the individual level. The providers had only one Excellent domain score: Avoidance of Substance Use (9.26/10). Nearly all providers had an Excellent score.

Individual Results

The LMA tool can also help to quickly identify concerning individual scores (individuals with low domain and/or overall scores). Our results revealed that most providers had substantial room for improvement; only 9/35 had an Excellent overall score ≥40. Further, most providers had at least one “lower” domain score; more specifically, 34/35 providers had at least one 7 or lower, 31/35 had at least one 6 or lower, and more than half (21/35) had a score of 4 or less in at least one domain. Only one participant had an excellent score (8 or higher) in five domains.

Results for Individual Questions

We analyzed the 21 LMA questions and revealed that the majority of the providers had poor scores for 3 questions and excellent scores for 6 questions. For the remaining 12 questions, the results were mixed, meaning high scores and low scores were both common. This suggests that scores tended to be unique to the individual. Results from the correlation matrix further corroborated this finding as there were very low correlations scores across the domains (r ≤ 0.35). Simply put, performance on one domain did not appear to be strongly related to performance on other domains. Overall, we determined that (1) most providers scored poorly on multiple questions, (2) most providers had at least one low domain score (area of concern), and (3) the low scores tend to be spread “randomly” throughout four of the five domains. In summary, the area of low scores tended to be unique to the individual. This suggests that individual health interventions would tend to be more useful/effective than group interventions.

Important Findings for Each Domain

Avoidance of Risky Substances

The providers excelled in this domain, with a group score of 9.26 out of 10. All 35 providers answered “no” to smoking, vaping, or e-cigarette use (6 points). The remaining two questions were pertaining to alcohol consumption with points awarded for 3 or less alcoholic drinks per week (4 or less if male) and one or less drinks per day (2 or less if male). For this domain, only four individuals scored less than 8 points.

Our cohort appears to be healthier (in this area) than typical medical providers previously reported in the literature.^33,35 Our cohort was in the family medicine specialty which could be a reason rates were lower. Previous work has revealed that specialties such as psychiatry, anesthesia, and emergency medicine are at higher risk and this may be due to the easy access to drugs, high-risk work environments and the baseline personalities of these healthcare providers. ³⁹ Occupational distress and job factors are known to increase the likelihood of providers using substances. ³⁶ Most of our providers reported managing stressors effectively most days, which may also help explain why the substance use rates were low.

Recovery

As previously mentioned, the Recovery domain score, which combines restorative sleep and stress management lifestyle pillars, appears to be the most concerning. The average sleep for providers was 6.3 ± 1.0 hours/night which resulted in a major deduction of points; the number of hours per night accounts for 5 of the 10 points. The full 5 points are only awarded if the provider sleeps 8+ hours of sleep per night. Three points are awarded for 7-8 hours of sleep. There were only 4 providers who reported 8+ hours, and 10 providers who reported 7 hours (awarded 3/5 points). Consequently, 24/35 providers reported 6 hours or less and received 0/5 points; nearly all had substantial room for improvement.

Limited sleep is common in physicians, especially for those who work on-call and night shifts.^40-42 Studies with high numbers of physicians have been completed. Karhula et al. ³⁶ studied over 700 hospital physicians in Finland and demonstrated that one fourth had ≤6.5 hours (self-reported). Wu et al. ²⁰ reported 6.37 hours of sleep for over 20,000 physicians in tertiary public hospitals in China. These results are similar to our cohort with 6.3 hours on average per night.

The providers in this study appear to cope well with limited sleep. The majority (54%) answered “yes” to waking up feeling refreshed and rested on most days and 89% felt they were able to manage and deal with stressors effectively most days. The fact that our cohort of providers are in a standard clinic setting (without night shifts) may contribute to the overall perception of feeling well-rested. Other studies have been completed for physicians who have regular night shifts. For example, high numbers of night work shifts (≥7 shifts/3 months) and higher total work week hours (>48 hours per week) were associated with higher odds for insufficient sleep.^40-42

Notably, 57% of our study respondents felt they did not have enough time to care for themselves most days. This may be a concern for burnout, which is common among physicians. A US study of 7000 physicians, 45.8% reported at least one symptom of burnout and rates were highest in specialties such as family medicine and internal medicine. Further, 40.1% of physicians reported limited time for personal or family life (23.1% for controls). This is related to physicians working above average hours; the median is 10 hours more per week compared to controls and a large proportion worked more than 60 hours per week. ⁴³ This likely explains why providers in this study reported limitations with self-care.

Movement

The providers in this study were highly variability in this area. The Movement domain has three questions. The highest-value question (5 points) addresses cardiorespiratory exercise. To obtain a score of 5/5, participants must engage in 2.5 hours or more of exercise per week, which aligns with the AHA and ACSM guidelines. There were 14 providers at or above guidelines. Ten were just under the recommended goal (1.5-2 hours) and 11 were well below (1.5 hours or less); this totals 21 or 60% below guidelines. These results are comparable to those reported by Aggarwal et al. ²⁸ who surveyed physician exercise habits. In that study, 49% of physicians were below the guidelines. Previous research has also revealed that physicians are not well-informed in this area. For example, only 46% of providers knew the AHA physical activity recommendations. ²⁸ Clearly, there is room for improvement in this domain.

For resistance training, two sessions weekly would score 2/2 as this is the target based on AHA and ACSM guidelines. Providers averaged 2.5 sessions, but with high variability: 17 met or exceeded guidelines, while 11 reported no resistance training. The third question evaluated daily sitting time, with 3 points awarded for 6 hours or less. The group averaged 5.7 ± 2.0 hours, with 23/35 staying below 6 hours daily. However, 7 individuals received zero points for sitting 8+ hours daily.

Nutrition

The providers had a mean Nutrition domain score of 7.34/10 which was categorized as “Very Good.” However, the results were again quite variable. This domain contains the most questions (6) on the survey: two 1-point questions and four 2-point questions. Our providers performed exceptionally well regarding the type of cooking oil (olive oil or no oil at all was the preferred to use), with 33/35 receiving a point for using olive oil as their primary cooking oil or using no oil. They also performed relatively well on limiting restaurant meals, with 25/35 providers earning the point for reporting three or fewer meals per week.

Fruit and vegetable consumption appears to be moderate, which is comparable to previous studies. In a Poland-based study, 64.4% of medical providers consumed three fruit portions per week but fewer than two portions daily. ²⁹ Our results showed 26/35 providers received the full 2 points for consuming 2+ servings daily of fruit, while 7 of the remaining 9 received partial credit (1 point) for consuming 1 serving daily. For vegetables, 19/35 received full points for consuming 3+ servings daily, with the remaining 16 receiving partial credit for 1-2 servings daily. The aforementioned study reported 80.7% of nurses and 56.5% of physicians consumed two to five servings of vegetables per week, but only 27.2% of the total sample met the threshold for more than five servings. ²⁸ This appears to be an area of need for providers to improve personal health and to better educate patients. ²⁸

Areas for potential improvement were also identified in the 2-point questions about sweetened drinks and packaged snacks. Nearly half (16/35) received zero points for consuming two or more sweetened drinks weekly, and 11/35 received zero points for consuming two or more packaged snacks daily. Wolska et al. ⁴⁴ previously analyzed unhealthy food choices among healthcare shift workers and similarly reported frequent snacking and high consumption of sugar-sweetened beverages. Those habits were particularly common among shift workers; this has been associated with disruptions in circadian rhythms, leading to increased cravings for high-calorie and processed foods. ⁴⁴ This is a contrast to our providers, who work in a clinical setting with regular hours, and may have more stable meal patterns. Our cohort appeared to reduce their reliance on convenience foods. However, both studies suggest substantial room for improvement (in multiple areas).

Connection

Connection domain results were also quite variable. This domain consists of five 2-point questions. The overall score was “Very Good,” (6.97/10) with 19/35 having a score ≥8. However, 16/35 had a concerning score of 6 or lower. As a group, the providers scored very well on two questions as 33/35 reported feeling their life had purpose, and 31/35 visited or spoke with close friends/family on three or more separate occasions. Therefore, deductions occurred mostly in the remaining three questions. Results were more mixed for nature contact (23/35 spent at least two hours in nature), spiritual/religious practices (20/35 engaged in two or more practices). The primary area for improvement was club/organization interactions, with more than half (20/35) answering “no” to interacting with one or more clubs or organizations. ³⁸

Previous studies have focused on loneliness in providers using tools including the UCLA Loneliness Scale and the Abbreviated Maslach Burnout Inventory (MBI-9). It has been reported that loneliness in practicing physicians is high and is associated with burnout and emotional exhaustion. ³⁸ The providers in this study appear to have a strong sense of purpose. However, there may be substantial room for improvement with social connectedness. Overall, it is unclear if our providers are suffering from loneliness, like the cohorts of physicians in previously reported studies. More in-depth assessments of loneliness would need to be completed to better understand. Previous work has highlighted the importance of creating opportunities to increase meaningful connections and to help find meaning in work. ³⁸

Conclusions

Lifestyle medicine is a medical specialty that is gaining popularity as an evidence-based practice to prevent, treat, and reverse chronic diseases that impacts millions of patients. The LMA tool can help to quickly identify domains that need improvement. Our results reveal that providers also have clear needs and substantial room for improvement across the domains. Further, it is now clear that there are important connections between the physicians’ health and the health outcomes and behaviors of their patients.^45,46

Overall, our results indicate substantial heterogeneity in domain scores for providers. The absence of strong correlations among domains underscores the likely need for individualized wellness approaches for healthcare providers.

This study has limitations that are important to note. The sample size was relatively small with 74% females. Also, we only assessed one cohort of family medicine providers at a single institution. The LMA tool itself is a brief questionnaire that only asks about behaviors in the last seven days.

Future studies using the LMA as a screening tool at other institutions and other specialties can help better understand the needs of medical providers.

Ethical considerations

This study was minimal risk and deemed IRB exempt

Consent to participate

Informed consent was provided in the email prior to the survey, and participants confirmed their consent by completing the survey questions.

ORCID iD

Johanna Mosquera-Moscoso https://orcid.org/0000-0002-8498-4910