Lifestyle patterns and sleep disturbance among Black breast cancer survivors: results from the Women’s Circle of Health Follow-Up Study

Abstract

Background

African American/Black breast cancer (BC) survivors experience higher rates of sleep disturbance than their White counterparts, which may contribute to worse survival and quality of life; however, whether adherence to cancer-related lifestyle guidelines (including adiposity, physical activity, and diet) is associated with post-diagnosis sleep disturbance in Black women remains unexplored. We evaluated whether adherence to World Cancer Research Fund/American Institute for Cancer Research (WCRF/AICR) recommendations was associated with a lower risk of sleep disturbance in a population-based cohort of Black BC survivors, the Women’s Circle of Health Follow-Up Study.

Methods

Pre-diagnosis lifestyle factors were assessed during home visits among 682 participants. Lifestyle pattern scores were calculated using the WCRF/AICR scoring system (range, 0-7; higher scores indicating greater adherence). Sleep disturbance was assessed ∼2 years post-diagnosis using the Pittsburgh Sleep Quality Index, with a score >5 indicating sleep disturbance. Robust Poisson regression estimated relative risks (RRs) and 95% confidence intervals (CIs) for associations between lifestyle pattern scores and sleep disturbance.

Results

The mean (SD) pre-diagnosis lifestyle pattern score was 3.0 (1.1), and 59.8% of survivors had post-diagnosis sleep disturbance. Compared with women in the lowest quartile of lifestyle pattern score (0-2.25), those in the highest quartile (≥4) had a 21% lower risk of sleep disturbance (RR, 0.79, 95% CI, 0.65, 0.97) after multivariable adjustment. Higher lifestyle pattern scores were associated with better domain scores for subjective sleep quality, sleep disturbance, and daytime dysfunction.

Conclusion

Adherence to cancer-related lifestyle guidelines may reduce the risk of sleep disturbance among Black BC survivors.

Implications for practice.

This study addresses an important gap by evaluating associations between adherence to cancer-related lifestyle guidelines and sleep disturbance among Black breast cancer survivors. Using a comprehensive lifestyle score, we evaluated the combined influence of adiposity, physical activity, and diet on sleep health. Higher adherence was associated with a lower risk of clinically significant sleep disturbance; however, average adherence remained low, and sleep disturbance 2 years post-diagnosis was highly prevalent. These findings suggest that cancer care should address sleep disturbance in this population and that multibehavior lifestyle interventions may improve post-diagnosis sleep quality.

Introduction

African American/Black women experience higher mortality, poorer quality of life, and greater symptom burden following a breast cancer diagnosis than White women.^1–4 Sleep is an important yet understudied dimension of cancer survivorship in this population. Sleep disturbance is highly prevalent among breast cancer survivors (BCS), with 38%-66% reporting poor sleep quality or insomnia symptoms.^5–12 In addition, Black BCS are more likely to experience poor sleep quality or shorter sleep duration compared to White BCS.^5,13 Sleep disturbance can negatively affect daily functioning and quality of life¹³ and may also contribute to long-term adverse outcomes following breast cancer, including cardiovascular disease, diabetes, and increased all-cause mortality.^5,7,14 Despite its potential relevance to both quality of life and long-term outcomes, few studies have specifically examined the role of sleep disturbance among Black BCS. This gap underscores the need to identify actionable strategies to improve sleep disturbance in this population.

The causes of disparities in sleep disturbance among Black BCS are multifactorial, going beyond differences in tumor biology and stage at diagnosis, and are likely influenced by modifiable lifestyle factors such as adiposity, physical activity, and dietary quality.^1,15 Maintaining a healthy body weight, consuming a nutritious diet, and engaging in regular physical activity have each been linked to reduced risk of cancer occurrence and recurrence, and improved overall survival.^16–18 Accordingly, the American Cancer Society and the World Cancer Research Fund/American Institute for Cancer Research (WCRF/AICR) recommend that cancer survivors also adopt these cancer-prevention behaviors to support long-term health.^16,19–21 Furthermore, pre-diagnosis lifestyle factors may reflect long-standing behaviors. Because health behaviors often co-occur, a lifestyle pattern approach that evaluates multiple behaviors simultaneously may better capture their combined and synergistic effects. However, little is known about the role of lifestyle factors in sleep disturbance among Black BCS,^22,23 and no prior study has examined the collective impact of these factors on sleep quality in this population. To address this gap, we examined the association between pre-diagnosis lifestyle patterns, evaluated through adherence to the WCRF/AICR cancer-related lifestyle recommendations, and sleep disturbance in a population-based cohort of Black BCS.

Methods

Study population and data collection

We used data from the Women’s Circle of Health Follow-Up Study (WCHFS), a longitudinal cohort of Black BCS that has been described in detail previously.²⁴ Participants were identified via rapid case ascertainment by the New Jersey State Cancer Registry (NJSCR) across 10 counties. Eligible women were aged 20-75 years at diagnosis, self-identified as Black or African American, had histologically confirmed ductal carcinoma in situ or invasive breast cancer, could speak English, and had no prior cancer history other than nonmelanoma skin cancer. All participants provided written informed consent, and the study was approved by Rutgers University and the Roswell Park Comprehensive Cancer Center Institutional Review Boards.

Baseline interviews were conducted approximately 10 months post-diagnosis, and Follow-Up 1 interviews about 24 months post-diagnosis. Both interviews were conducted during home visits and administered using computer-assisted questionnaires that captured information on sociodemographic characteristics, medical and reproductive history, lifestyle factors, as well as patient-reported outcomes. Anthropometric measurements and blood pressure assessments were also obtained during the visits.

Clinicopathologic characteristics, including tumor stage, hormone receptor status, and human epidermal growth factor receptor 2 status, were obtained from pathology reports and NJSCR data linkages. BC treatment information was abstracted from medical records and supplemented with NJSCR data and self-report; self-reported treatment data showed high concordance with medical records.²⁵ History of hypertension and diabetes was obtained from medical records and self-reports, with hypertension also identified based on blood pressure measurements during home visits.²⁵

Among 800 women diagnosed between 2012 and 2019 who completed both the baseline and Follow-Up 1 visits, 18 were excluded due to missing sleep disturbance data, 97 due to missing lifestyle information (primarily noncompletion of the food frequency questionnaire [FFQ]), 2 due to extreme energy intake (greater than twice the interquartile range of log energy intake),^26,27 and 1 due to missing covariate data. Consequently, 682 women were included in the final analysis.

Lifestyle assessment and scoring

Trained interviewers recorded anthropometric measurements, including weight, height, and waist circumference, following standardized protocols.^24,25 Baseline measurements were used for analyses; when measured weight or height was missing, self-reported values were used, which have demonstrated high concordance with measured values in this cohort.²⁸ Physical activity during the year before diagnosis was evaluated using the Black Women’s Health Study (BWHS) Physical Activity Questionnaire,²⁹ and dietary intake during the same period was assessed via the validated FFQ adapted from the BWHS.^30,31

Adherence to cancer-related lifestyle recommendations was evaluated using the 2018 WCRF/AICR score.^32,33 The score operationalizes the 2018 WCRF/AICR Cancer Prevention Recommendations into 7 components reflecting adiposity, physical activity, and dietary behaviors: (1) maintain a healthy weight; (2) be physically active; (3) consume a diet rich in fruits, vegetables, and fibers; (4) limit intake of ultra-processed foods high in fat, starches, or sugars; (5) reduce consumption of red and processed meats; (6) minimize intake of sugar-sweetened beverages; and (7) limit alcohol consumption. Each component of the WCRF/AICR score was assigned points based on the level of adherence to the recommendation using standardized criteria. The weight and fruit/vegetable/fiber components were scored in 0.25-point increments, and others in 0.5-point increments. Component scores were summed to generate a total lifestyle score ranging from 0 to 7 points, with higher scores reflecting greater adherence to recommendations (scoring criteria are detailed in Table S1).

Sleep disturbance outcome measurement

Sleep disturbance over the past month was assessed at the Follow-Up 1 home visit (∼24 months after breast cancer diagnosis) using the Pittsburgh Sleep Quality Index (PSQI).^5,34 This 18-item questionnaire evaluates 7 domains: subjective sleep quality, sleep latency (time to fall asleep), sleep duration, sleep efficiency (percent of time in bed that is spent asleep), sleep disturbance (frequency of night awakenings), use of sleeping medication, and daytime dysfunction (difficulty staying awake during the day). Each domain is scored from 0 to 3, yielding a total PSQI score ranging from 0 to 21. A global score ≤5 indicates good sleep quality, while a score >5 suggests clinically significant sleep disturbance.³⁴ The PSQI has demonstrated strong reliability and validity in cancer patients and survivorship populations.^35,36

Statistical analysis

Participant characteristics were compared between participants with clinically significant sleep disturbance and those without sleep disturbance (PSQI >5 vs ≤5) using t-tests or chi-square tests, as appropriate.

Poisson regression with robust standard errors was used to estimate relative risks (RRs) and 95% confidence intervals (CIs) for associations between lifestyle pattern scores and sleep disturbance. Lifestyle patterns were evaluated both as a continuous score and in quartiles. Quartiles were defined as follows: Q1 (score 0-2.25), representing the lowest adherence to the lifestyle recommendations; Q2 (score 2.5-3); Q3 (score 3.25-3.75); and Q4 (score ≥4), representing the highest adherence. For associations between lifestyle pattern scores and PSQI component scores, linear regression was used to estimate β coefficients and 95% CIs.

Age-adjusted and multivariable-adjusted models were fitted. Covariates included in the multivariable-adjusted model were selected a priori based on existing literature and a directed acyclic graph and included age at diagnosis, education level, insurance status, household income, and smoking status (Figure S1). Other variables, including menopausal status, marital status, tumor stage, subtype, breast cancer treatment, comorbidities (diabetes and hypertension), depression, and anxiety, were also evaluated in the directed acyclic graph but were not identified as confounders. Sensitivity analyses additionally adjusted for tumor stage and subtype, cancer treatment, comorbidities, marital status, menopausal status, and neighborhood socioeconomic status,³⁷ or excluded participants with depression or anxiety.

Potential effect modification by age at diagnosis, menopausal status, tumor subtype, tumor stage, smoking status, endocrine therapy, and chemotherapy was evaluated. P value for multiplicative interaction were estimated using likelihood ratio tests.

To evaluate the contribution of each lifestyle component to the overall association with sleep disturbance, we examined each component individually while adjusting for the overall lifestyle pattern score excluding the component under consideration.

All tests were two-sided, with statistical significance defined as P < .05. All statistical analyses were conducted using Stata version 18.5 (StataCorp LLC, TX, USA).

Results

Among the 682 Black breast cancer survivors included in the analysis, the mean (±SD) age at diagnosis was 55.1 (±10.7) years (Table 1). The mean (±SD) PSQI score approximately 2 years post-diagnosis was 7.4 (±4.6) points, and 408 participants (59.8%) met criteria for clinically significant sleep disturbance. Compared with women without sleep disturbance, those with sleep disturbance were more likely to be younger at diagnosis, premenopausal, lack health insurance, and to have received chemotherapy. The mean (±SD) pre-diagnosis lifestyle pattern score was 3.0 (±1.1). Fewer than half of participants met each individual recommended lifestyle guideline for BMI and waist circumference, physical activity, red and processed meat intake, or sugar-sweetened beverage consumption (Table S1).

Table 1.

Distribution of participant characteristics overall and by sleep disturbance status in the Women’s Circle of Health Follow-Up Study.

Characteristics	Total (N = 682)	Clinically significant sleep disturbance
		No (N = 274)	Yes (N = 408)	P valuea
Age at diagnosis, mean (SD)	55.1 (10.7)	56.3 (11.0)	54.2 (10.4)	.012
Education
High school or less	236 (34.6%)	87 (31.8%)	149 (36.5%)	.079
Some college	228 (33.4%)	86 (31.4%)	142 (34.8%)
College or postgraduate	218 (32.0%)	101 (36.9%)	117 (28.7%)
Household income				.23
<$25 000	205 (30.1%)	71 (25.9%)	134 (32.8%)
$25 000-$69 999	229 (33.6%)	95 (34.7%)	134 (32.8%)
≥$70 000	217 (31.8%)	96 (35.0%)	121 (29.7%)
Unknown	31 (4.5%)	12 (4.4%)	19 (4.7%)
Marital status				.068
Married/living as married	244 (35.8%)	112 (40.9%)	132 (32.4%)
Single/never married	217 (31.8%)	80 (29.2%)	137 (33.6%)
Divorced/separated	163 (23.9%)	56 (20.4%)	107 (26.2%)
Widow	58 (8.5%)	26 (9.5%)	32 (7.8%)
Insurance status				.015
Private	376 (55.1%)	154 (56.2%)	222 (54.4%)
Medicaid	88 (12.9%)	29 (10.6%)	59 (14.5%)
Medicare	125 (18.3%)	57 (20.8%)	68 (16.7%)
Other/Unknown	34 (5.0%)	19 (6.9%)	15 (3.7%)
Uninsured	59 (8.7%)	15 (5.5%)	44 (10.8%)
Menopausal status				.035
Pre-menopausal	236 (34.6%)	82 (29.9%)	154 (37.7%)
Post-menopausal	446 (65.4%)	192 (70.1%)	254 (62.3%)
Smoking status				.30
Never smoker	413 (60.6%)	164 (59.9%)	249 (61.0%)
Former smoker	164 (24.0%)	73 (26.6%)	91 (22.3%)
Current smoker	105 (15.4%)	37 (13.5%)	68 (16.7%)
Diabetes				.59
No	498 (73.0%)	197 (71.9%)	301 (73.8%)
Yes	184 (27.0%)	77 (28.1%)	107 (26.2%)
Hypertension				.48
No	243 (35.6%)	102 (37.2%)	141 (34.6%)
Yes	439 (64.4%)	172 (62.8%)	267 (65.4%)
Tumor stage				.097
0	132 (19.5%)	62 (22.9%)	70 (17.3%)
I	255 (37.7%)	108 (39.9%)	147 (36.3%)
II	223 (33.0%)	78 (28.8%)	145 (35.8%)
III & IV	66 (9.8%)	23 (8.5%)	43 (10.6%)
Unknown	6	3	3
Tumor subtype				.30
Luminal A	342 (50.1%)	130 (47.4%)	212 (52.0%)
HER2+	128 (18.8%)	48 (17.5%)	80 (19.6%)
TNBC	117 (17.2%)	51 (18.6%)	66 (16.2%)
Unknown	95 (13.9%)	45 (16.4%)	50 (12.3%)
Type of surgery				.057
No surgery	21 (3.1%)	10 (3.6%)	11 (2.7%)
Lumpectomy	347 (50.9%)	153 (55.8%)	194 (47.5%)
Mastectomy	314 (46.0%)	111 (40.5%)	203 (49.8%)
Chemotherapy				.048
No	317 (46.5%)	140 (51.1%)	177 (43.4%)
Yes	365 (53.5%)	134 (48.9%)	231 (56.6%)
Hormone therapy				.42
No	212 (31.1%)	90 (32.8%)	122 (29.9%)
Yes	470 (68.9%)	184 (67.2%)	286 (70.1%)
Radiation therapy				.75
No	192 (28.2%)	79 (28.8%)	113 (27.7%)
Yes	490 (71.8%)	195 (71.2%)	295 (72.3%)
PSQI score, mean (SD)	7.4 (4.6)	3.1 (1.4)	10.4 (3.5)	<.001
Sleep efficiency %, mean (SD)	78.4 (19.2)	90.5 (9.3)	70.2 (19.9)	<.001
Hours of sleep per night, mean (SD)	5.9 (1.5)	6.9 (1.0)	5.3 (1.4)	<.001
Lifestyle pattern score, mean (SD)b	3.0 (1.1)	3.1 (1.1)	2.9 (1.1)	.005
Q1: Score 0-2.25	205	76	129	.043
Q2: Score 2.5-3	142	53	89
Q3: Score 3.25-3.75	195	74	121
Q4: Score ≥4	140	71	69

Abbreviations: PSQI, Pittsburgh Sleep Quality Index; HER2+, human epidermal growth factor receptor 2; TNBC, triple-negative breast cancer.

^a P value by t-tests or chi-square tests, as appropriate (unknowns excluded for stage).

^bThe lifestyle pattern score ranges from 0 to 7, in increments of 0.25, with a higher score indicating better adherence to the guidelines.

Using Poisson regression models with robust standard errors, both the age-adjusted and multivariable-adjusted models showed consistent findings (Table 2). In the multivariable model, participants in the highest quartile of lifestyle pattern scores (Q4: ≥4) had a 21% lower risk of sleep disturbance (RR, 0.79, 95% CI, 0.65, 0.97) compared to those in the lowest quartile (Q1: 0-2.25). No significant associations were observed for Q2 (score 2.5-3) or Q3 (score 3.25-3.75) compared with Q1. When the lifestyle score was treated as a continuous variable, each one-point increase in the lifestyle pattern score was associated with a 7% lower risk of sleep disturbance (RR, 0.93, 95% CI, 0.87, 0.98).

Table 2.

Associations between lifestyle pattern scores and sleep disturbance among Black breast cancer survivors.^a

Lifestyle pattern scoreb	Participants, N	RR	(95% CI)
Model 1 (age-adjusted model)
Q1: Score 0-2.25 low adherence	205	Reference
Q2: Score 2.5-3	142	1.00	(0.85, 1.18)
Q3: Score 3.25-3.75	195	1.00	(0.86, 1.17)
Q4: Score ≥4 high adherence	140	0.79	(0.65, 0.97)
Per 1-score increase	682	0.93	(0.88, 0.98)
Model 2 (multivariable-adjusted model)c
Q1: Score 0-2.25 low adherence	205	Reference
Q2: Score 2.5-3	142	1.01	(0.85, 1.19)
Q3: Score 3.25-3.75	195	1.00	(0.86, 1.17)
Q4: Score ≥4 high adherence	140	0.79	(0.65, 0.97)
Per 1-score increase	682	0.93	(0.87, 0.98)

Abbreviations: CI, confidence interval; RR, relative risk.

^aSleep disturbance is defined as PSQI > 5. The associations were estimated via Poisson regression with robust standard errors.

^bThe lifestyle pattern score ranges from 0 to 7, in increments of 0.25, with higher scores indicating better guideline adherence.

^cModel 2 adjusted for age at diagnosis, education, household income, insurance status and smoking status.

Sensitivity analyses further adjusting for tumor stage and subtype showed similar findings, with a persistent inverse association observed for participants in the highest lifestyle score quartile (RR, 0.79, 95% CI, 0.65, 0.98; Table S2). Results also remained consistent when analyses were additionally adjusted for neighborhood socioeconomic status; comorbidities, menopausal status, and marital status; or cancer treatment, as well as after restricting analyses to participants without depression or anxiety (Table S2). In a post hoc analysis using Q2 or Q3 as the reference group, participants in Q4 continued to demonstrate significantly lower risk of sleep disturbance, supporting a threshold effect at higher levels of adherence to lifestyle recommendations (Table S3).

No statistically significant effect modification was observed by age at diagnosis, menopausal status, smoking status, tumor stage, tumor subtype, or receipt of endocrine therapy or chemotherapy (P for interaction=.20-.87; Table S4), indicating that the inverse association between lifestyle patterns and sleep disturbance was consistent across these subgroups.

In analyses examining the individual components of the WCRF/AICR lifestyle score, most components were not significantly associated with sleep disturbance (Table S5). Only higher fruit and vegetable intake (≥400 g/day vs <200 g/day) was borderline-significantly associated with a lower risk of sleep disturbance (RR, 0.85, 95% CI, 0.73-0.99).

Associations between lifestyle pattern scores and individual PSQI components were also examined (Table 3). In multivariable linear regression models, participants in the highest lifestyle pattern adherence category (Q4) had more favorable scores for selected sleep domains compared with those in the lowest category (Q1). Specifically, Q4 was associated with a borderline-significant 0.21-point lower score (95% CI, −0.41 to 0.00) for subjective sleep quality, a 0.18-point lower score (95% CI, −0.32 to −0.03) for the sleep disturbance component, and a 0.19-point lower score (95% CI, −0.36 to −0.02) for daytime dysfunction. No significant associations were observed for the other sleep components.

Table 3.

Associations between lifestyle pattern scores and PSQI subscales among Black breast cancer survivors.^a

Lifestyle pattern scoreb		Model 1 (age-adjusted model)		Model 2 (multivariable-adjusted model)c
	N	Coefficient	95% CI	Coefficient	95% CI
Global PSQI (score range: 0-21)
Q1: Score 0-2.25	197	Reference		Reference
Q2: Score 2.5-3	135	−0.36	(−1.35, 0.64)	−0.26	(−1.25, 0.73)
Q3: Score 3.25-3.75	192	−0.20	(−1.11, 0.70)	−0.04	(−0.95, 0.88)
Q4: Score ≥4	140	−1.48	(−2.46, −0.49)	−1.20	(−2.23, −0.18)
Per 1-score increase	664	−0.48	(−0.79, −0.16)	−0.39	(−0.72, −0.06)
PSQI subscales (score range: 0-3 for each domain)
Subjective sleep quality
Q1: Score 0-2.25	197	Reference		Reference
Q2: Score 2.5-3	135	0.04	(−0.16, 0.24)	0.05	(−0.15, 0.25)
Q3: Score 3.25-3.75	192	0.03	(−0.15, 0.21)	0.02	(−0.16, 0.21)
Q4: Score ≥4	140	−0.20	(−0.40, −0.01)	−0.21	(−0.41, −0.00)
Per 1-score increase	664	−0.07	(−0.13, −0.01)	−0.07	(−0.14, −0.01)
Sleep latency
Q1: Score 0-2.25	197	Reference		Reference
Q2: Score 2.5-3	135	−0.04	(−0.27, 0.19)	−0.03	(−0.26, 0.20)
Q3: Score 3.25-3.75	192	0.02	(−0.19, 0.23)	0.04	(−0.17, 0.26)
Q4: Score ≥4	140	−0.19	(−0.42, 0.04)	−0.17	(−0.41, 0.07)
Per 1-score increase	664	−0.06	(−0.13, 0.01)	−0.05	(−0.12, 0.03)
Sleep duration
Q1: Score 0-2.25	197	Reference		Reference
Q2: Score 2.5-3	135	0.02	(−0.22, 0.27)	0.03	(−0.22, 0.27)
Q3: Score 3.25-3.75	192	−0.01	(−0.23, 0.21)	0.02	(−0.20, 0.24)
Q4: Score ≥4	140	−0.14	(−0.38, 0.10)	−0.09	(−0.34, 0.16)
Per 1-score increase	664	−0.06	(−0.14, 0.02)	−0.04	(−0.12, 0.04)
Sleep efficiency
Q1: Score 0-2.25	197	Reference		Reference
Q2: Score 2.5-3	135	−0.12	(−0.38, 0.15)	−0.11	(−0.37, 0.16)
Q3: Score 3.25-3.75	192	−0.09	(−0.33, 0.15)	−0.05	(−0.29, 0.19)
Q4: Score ≥4	140	−0.28	(−0.54, −0.02)	−0.20	(−0.48, 0.07)
Per 1-score increase	664	−0.08	(−0.17, −0.00)	−0.06	(−0.15, 0.03)
Sleep disturbance
Q1: Score 0-2.25	197	Reference		Reference
Q2: Score 2.5-3	135	−0.08	(−0.23, 0.06)	−0.07	(−0.21, 0.07)
Q3: Score 3.25-3.75	192	−0.03	(−0.16, 0.10)	−0.01	(−0.15, 0.12)
Q4: Score ≥4	140	−0.19	(−0.33, −0.05)	−0.18	(−0.32, −0.03)
Per 1-score increase	664	−0.05	(−0.10, −0.01)	−0.05	(−0.09, 0.00)
Use of sleep medication
Q1: Score 0-2.25	197	Reference		Reference
Q2: Score 2.5-3	135	−0.14	(−0.38, 0.10)	−0.11	(−0.35, 0.13)
Q3: Score 3.25-3.75	192	−0.02	(−0.23, 0.20)	0.02	(−0.20, 0.24)
Q4: Score ≥4	140	−0.27	(−0.51, −0.03)	−0.17	(−0.42, 0.08)
Per 1-score increase	664	−0.08	(−0.16, −0.01)	−0.05	(−0.13, 0.03)
Daytime dysfunction
Q1: Score 0-2.25	197	Reference		Reference
Q2: Score 2.5-3	135	−0.04	(−0.20, 0.12)	−0.03	(−0.19, 0.13)
Q3: Score 3.25-3.75	192	−0.10	(−0.25, 0.05)	−0.09	(−0.23, 0.06)
Q4: Score ≥4	140	−0.20	(−0.36, −0.04)	−0.19	(−0.36, −0.02)
Per 1-score increase	664	−0.07	(−0.12, −0.02)	−0.07	(−0.12, −0.01)

Abbreviations: PSQI, Pittsburgh Sleep Quality Index; CI, confidence interval.

^aParticipants with missing data on any subscale were excluded from this analysis. The associations were estimated via linear regressions.

^bThe lifestyle pattern score ranges from 0 to 7, in increments of 0.25, with higher scores indicating better guideline adherence.

^cModel 2 adjusted for age at diagnosis, education, household income, insurance status and smoking status.

Discussion

In this large, population-based cohort of Black BCS, a healthier pre-diagnosis lifestyle pattern, as evaluated by the 2018 WCRF/AICR score, was associated with a significantly lower risk of sleep disturbance. This reduced risk was only observed among participants in the highest quartile of lifestyle pattern scores (≥4 out of 7) compared with those in the lowest or middle quartiles, suggesting that greater adherence to multiple modifiable lifestyle behaviors may be required to improve sleep outcomes among BCS. Beyond the association with global sleep quality, participants in the highest adherence category also demonstrated more favorable scores for specific sleep domains, including subjective sleep quality, sleep disturbance, and daytime dysfunction.

Notably, 59.8% of Black BCS in this study experienced clinically significant sleep disturbance approximately 2 years post-diagnosis. We have previously documented that Black BCS may experience persistent sleep disturbance from before diagnosis throughout survivorship.⁵ Except for our previous report based on a smaller sample,⁵ few studies have examined sleep disturbance in a large population-based cohort of Black BCS. The prevalence we observed among Black BCS is considerably higher than the pooled estimate of 40% (95% CI, 0.29-0.52) reported in a meta-analysis among BCS,⁷ and substantially higher than the 30%-40% prevalence reported among Black women without a cancer diagnosis.^38,39

Most prior research on sleep in cancer survivorship has focused on the influence of individual lifestyle factors, such as BMI,^40–42 physical activity,^43–46 or diet,^47,48 typically examined in isolation. A UK-based cross-sectional study examining adherence to individual WCRF/AICR recommendations among breast, prostate, and colorectal cancer survivors found no association between diet, physical activity, or alcohol consumption and sleep quality, whereas obesity was associated with poorer sleep.⁴⁰ While these studies provide important insights, single-factor approaches may not adequately capture the overall influence of co-occurring health behaviors. Supporting this perspective, an analysis of NHANES data found that the combination of improved dietary quality and physical activity was associated with lower odds of unhealthy sleep patterns among cancer survivors.⁴⁹ The present study extends this research by evaluating overall lifestyle patterns using a standardized composite score in relation to sleep disturbance after breast cancer diagnosis. In our analyses of individual WCRF/AICR components, most factors were not independently associated with sleep disturbance, except for higher fruit and vegetable intake, which showed a modest inverse association. These findings suggest that no single lifestyle behavior predominates in influencing sleep disturbance and support the use of a composite lifestyle approach to better capture the combined potential effects of multiple behaviors among breast cancer survivors.

These findings are particularly relevant to Black BCS, who experience disproportionately higher rates of sleep disturbance and adverse breast cancer outcomes compared with other racial and ethnic groups.^5,24,50–54 Although a reduced risk of sleep disturbance was observed among Black BCS with high pre-diagnosis adherence to the 2018 WCRF/AICR cancer-prevention lifestyle recommendations, the mean (±SD) score in this cohort was only 3.0 (±1.1). By comparison, cancer-free, predominantly White women in the NIH-AARP Diet and Health Study had a mean (±SD) score of 3.5 ± 1.0 using the same criteria.⁵⁵ Differences in cancer status, as well as structural and social determinants of health, including income, education, neighborhood environment, cultural norms, and access to health care, may contribute to the lower lifestyle adherence observed in this population.^56,57

Most participants did not meet lifestyle recommendations for several key components, including body weight, physical activity, red and processed meat intake, and sugar-sweetened beverage consumption. These domains represent potentially important targets for intervention in this population. However, it is essential to recognize that adoption and maintenance of healthy lifestyle behaviors are not equally feasible for all individuals, as healthier behaviors are strongly patterned by socioeconomic status.^58,59 Accordingly, efforts to improve adherence to cancer-related lifestyle guidelines should incorporate strategies that address structural and social determinants of health in order to promote equitable opportunities for adopting healthy behaviors.

Higher adherence to cancer-related lifestyle guidelines appeared to be associated with specific sleep domains, including subjective sleep quality, sleep disturbance, and daytime dysfunction, but not with other domains, including sleep duration, sleep latency, sleep efficiency, or use of sleep medication. It is possible that some sleep components are more strongly influenced by non-lifestyle factors, such as cancer treatment, pain, or medication use. In addition, the restricted 0-3 range of the domain-specific scores may have reduced our ability to detect associations across some sleep components.

Lifestyle patterns may influence sleep through multiple interrelated pathways, including inflammation, hormonal signaling, circadian rhythm alignment, and metabolic regulation.^60–66 Regular physical activity supports circadian stability and reduces pro-inflammatory cytokines.^63,65,66 Diets high in fruits and vegetables promote anti-inflammatory balance, whereas higher intake of red and processed meats and sugar-sweetened beverages contributes to chronic inflammation and metabolic dysfunction that may impair sleep.^60,62 Excess adiposity may further disrupt sleep through increased systemic inflammation, insulin resistance, and mechanical disturbances such as obstructive sleep apnea.⁶⁴ Alcohol consumption also interferes with circadian hormonal signaling by suppressing melatonin and provoking inflammation.⁶¹ Furthermore, healthier pre-diagnosis lifestyle patterns may contribute to better nutritional status and be associated with preserved muscle mass during treatment.²⁶ This, in turn, may reduce treatment-related toxic effects and improve patient-reported outcomes, including sleep.

These findings have implications for intervention development. Given that the observed potential benefit was limited to those with the highest adherence to multiple lifestyle recommendations, multibehavior interventions that concurrently target diet, physical activity, and weight management may be needed to meaningfully improve sleep outcomes among breast cancer survivors. Potential strategies include structured programs that integrate weight management, dietary counseling, and supervised or home-based physical activity delivered through multidisciplinary teams. Prior studies support the feasibility of combined diet and physical activity interventions among female cancer survivors.^17,45,46 To promote implementation, strategies to mitigate structural barriers to multibehavior interventions should be considered. These may include community-based delivery models and culturally tailored programming to improve relevance and engagement.^56,57 Flexible or home-based formats, and partnerships with local organizations may further enhance accessibility.

Limitations of the study should be acknowledged. Several lifestyle factors (diet and physical activity) and sleep quality were self-reported. Although validated questionnaires were used, measurement error is possible. Objective sleep measures (such as actigraphy) and objective assessments of physical activity using wearable devices were not feasible in this population-based cohort; future longitudinal studies incorporating these tools may help validate and extend these findings. Because our analysis was restricted to breast cancer survivors, results may also be influenced by index event bias, as potential shared determinants (such as pre-diagnosis psychosocial stress) may affect both cancer risk and post-diagnosis sleep outcomes.^67,68 In addition, our exposure reflects pre-diagnosis lifestyle patterns, which may not fully capture behaviors after a breast cancer diagnosis. Prior studies suggest that while some survivors report modest improvements in lifestyle behaviors, these changes are variable and not usually sustained.^69,70 Future studies should also evaluate post-diagnosis lifestyle patterns to better inform modifiable targets for intervention. Finally, despite adjusting for a wide array of covariates and conducting sensitivity analyses that support the robustness of our findings, residual confounding cannot be ruled out.

This study has several strengths. To our knowledge, it is the first investigation to specifically examine lifestyle patterns and sleep disturbance in a large, population-based cohort of Black BCS. Another strength is the lifestyle pattern approach, which captures multiple health behaviors simultaneously. This integrative approach reflects the real-world scenario, in which adiposity, physical activity, dietary quality, and alcohol consumption interact to influence health outcomes. It also enabled assessment of the potential combined effect of these behaviors on sleep disturbance, rather than evaluating each behavior in isolation. Additionally, data in WCHFS were collected through in-person interviews and standardized questionnaires, complemented by other data sources, allowing for careful consideration of potential confounders and effect modifiers in the analysis.

In conclusion, a healthier overall lifestyle pattern was associated with a lower risk of sleep disturbance among Black BCS, with the association evident among those with the highest overall adherence to lifestyle recommendations. These findings underscore the high burden of sleep disturbance in this population and suggest that simultaneous adherence to multiple healthy lifestyle behaviors may be particularly important for improving sleep quality.

Author contributions

Amogh Kudva (Conceptualization, Formal analysis, Writing—original draft, Writing—review & editing), Elisa V. Bandera (Conceptualization, Data curation, Funding acquisition, Writing—review & editing), Shromona Sarkar (Formal analysis, Writing—review & editing), Nur Zeinomar (Conceptualization, Data curation, Writing—review & editing), Tengteng Wang (Data curation, Writing—review & editing), Brian D. Gonzalez (Data curation, Writing—review & editing), Coral O. Omene (Writing—review & editing), Mi-Hyeon Jang (Writing—review & editing), Kitaw Demissie (Data curation, Funding acquisition, Writing—review & editing), Chi-Chen Hong (Data curation, Funding acquisition, Writing—review & editing), and Bo Qin (Conceptualization, Data curation, Funding acquisition, Supervision, Writing—review & editing)

Supplementary material

Supplementary material is available at The Oncologist online.

Funding

This work was supported by grants from the American Cancer Society RSG-23-1143513-01-CTPS (Qin), Rutgers Robert Wood Johnson Medical School EJI Excellence in Medicine Endowment Fund (Kudva), the National Cancer Institute R01CA185623 (Bandera, Demissie, and Hong), P30CA072720-5929 (Bandera), and the Breast Cancer Research Foundation (Ambrosone and Hong). The New Jersey State Cancer Registry, Cancer Epidemiology Services, New Jersey Department of Health, is funded by the Surveillance, Epidemiology and End Results (SEER) Program of the National Cancer Institute under contract 75N91021D00009, the National Program of Cancer Registries (NPCR), Centers for Disease Control and Prevention under grant NU58DP007117 as well as the State of New Jersey and the Rutgers Cancer Institute.

Conflicts of interest

E.V.B. is supported by the Rutgers Unilever Endowed Chair in Nutrition and Cancer Prevention and served on a Pfizer Inc. Advisory Board to enhance minority participation in clinical trials, both unrelated to this work. B.D.G. receives support from Janssen and Elly Health unrelated to this work. All other authors report no disclosures pertinent to the submitted work.

Data availability

The de-identified data underlying this article are available upon approval of a data request form by the Women’s Circle of Health Follow-up Study Scientific Committee and contingent upon appropriate human subjects research approval and a data transfer agreement. For further information, please contact the corresponding author.