Biological Sex Differences in Risk Factors and Outcomes Among Hospitalized Adults With Stroke in Lusaka, Zambia

Abstract

Objective

We investigated sex differences in clinical characteristics and outcomes among hospitalized adults with stroke in Zambia.

Methods

We retrospectively collected information for 324 consecutively hospitalized adults with stroke on the neurology service at the University Teaching Hospital in Lusaka, Zambia, between October 2018 and March 2019. Stroke characteristics were then compared by biological sex.

Results

Female participants constituted 62% (n = 200) of the cohort, were older (61 ± 19 vs 57 ± 16 years, p = 0.06), had fewer hemorrhagic stroke than male participants (22% vs 37%, p = 0.001), and had higher rates of hypertension (84% vs 74%, p = 0.04), diabetes (19% vs 13%, p = 0.04), heart disease (38% vs 27%, p = 0.04), and history of stroke (26% vs 14%, p = 0.01). Male participants had higher rates of alcohol (33% vs 4%, p < 0.001) and tobacco (19% vs 2%, p < 0.001) use. Female participants were less likely to have neuroimaging completed during their hospitalization (82% vs 94%, p = 0.002) and had higher 90 days postdischarge mortality (28% vs 10%, p = 0.002) independent of age and stroke subtype (OR 2.48, 95% CI 1.1–5.58, p = 0.03).

Discussion

Female participants in this Zambian stroke cohort had a higher prevalence of vascular risk factors but were less likely to have neuroimaging completed. Postdischarge mortality was markedly higher among female participants even after adjusting for age and stroke subtype. Our data highlight the need for future studies of social and socioeconomic factors that may influence stroke-related outcomes.

While stroke subtypes and risk factors differ by sex in some populations, little is known about sex differences in sub-Saharan Africa. Here, we describe sex differences in stroke risk factors, subtypes, care, and mortality among adults hospitalized with stroke in Zambia.

Methods

Setting

This study was conducted at the University Teaching Hospital (UTH), Zambia's national referral hospital located in the capital city of Lusaka. Inpatients at UTH do not pay admission fees, physician fee, or for basic laboratory investigations or medications available in the hospital pharmacy but are required to pay for CT scans.

Study Population

A retrospective cohort study was conducted for adults admitted to UTH with stroke between October 2018 and March 2019. Detailed study methods, definitions of study variables, and characteristics of this cohort have been published previously.^1,2

Study Procedures

We abstracted demographics, clinical presentation, comorbidities, examination findings, laboratory and imaging results, in-hospital complications, and vital status at discharge from paper charts. Biological sex was assigned by the admitting physician. Stroke was classified as ischemic or hemorrhagic based on CT imaging and “unknown” if stroke was clinically suspected, but no imaging was performed. Participants' vital status at 90 days after discharge was obtained from outpatient neurology appointment attendance and phone calls. Ninety days postdischarge mortality was defined as individuals who survived their hospitalization but died in the first 90 days after discharge.

Statistical Analyses

Descriptive statistics were compared by sex. Multivariable logistic regression models including age, sex, and stroke subtype assessed the independent contribution of sex to in-hospital and 90 days postdischarge mortality. Stata 14 (College Station, TX) was used.³

Standard Protocol Approvals, Registrations, and Patient Consents

This study did not collect identifying patient information, so informed consent was not obtained. The University of Zambia Biomedical Research Ethics Council, the Zambia National Health Research Authority, and the Johns Hopkins Institutional Review Board approved this study.

Data Availability

Deidentified study data are available on reasonable request after receiving necessary ethics and regulatory approvals in Zambia.

Results

In this cohort of 324 adults hospitalized with stroke, 62% (n = 200) were female. Female participants were older (61 ± 19 years vs 57 ± 16 years (male participants) (p = 0.06) and had fewer hemorrhagic strokes (22% vs 37%, p = 0.001) and a higher frequency of unknown strokes (18% vs 6%, p = 0.001) than male participants (Table 1). Among participants with ischemic stroke, female participants were more likely to have a stroke suspected to be from cardioembolism (30% vs 11% in male participants, p = 0.003). Compared with male participants, female participants also had higher rates of several stroke risk factors, including hypertension (84% vs 74%, p = 0.04), diabetes (19% vs 10%, p = 0.04), and history of stroke (26% vs 14%, p = 0.01). Significantly fewer female participants (82%) than male participants (94%) obtained head CT scans during admission at UTH (p = 0.002) (Table 1). In-hospital mortality did not vary significantly by sex (27% in female participants vs 22% in male participants, p = 0.3) (Figure). Female participants had a significantly higher 90 days postdischarge mortality than male participants (28% vs 10%, p = 0.002) even after adjusting for age and stroke subtype (OR 2.48, 95% CI 1.1–5.58, p = 0.03) (Table 2).

Table 1.

Differences in Stroke Demographics, Clinical Characteristics, and Risk Factors Between Male and Female Participants

Figure. In-Hospital Mortality Among Adults Admitted With Stroke and 90 Days Postdischarge Mortality Among Adults Who Survived Their Initial Hospitalization Compared by Sex.

Table 2.

Results of Multivariable Logistic Regression Models Assessing the Contribution of Sex to In-Hospital and 90 Days Postdischarge Mortality After Adjusting for Age and Stroke Subtype

Discussion

Female participants in our stroke cohort were more likely to have ischemic strokes, potentially because of older age, which increases ischemic stroke risk,⁴ and loss of estrogen neuroprotection against brain ischemia before menopause.⁵ Female participants also had higher rates of cardioembolic strokes and ischemic stroke risk factors such as hypertension and prior stroke.

Notably, female participants in our cohort were significantly less likely to have neuroimaging performed and experienced markedly higher 90 days postdischarge mortality. Together, these data suggest differences in in-hospital and poststroke systems of care delivery based on sex and may indicate that female participants in this urban Zambian cohort experience systemic obstacles in accessing stroke care. As male participants are regarded as the primary wage earners in the family, whereas female participants take on caregiver roles in Zambia⁶; female individuals may have less access to procure funds for medical care as family members may be less likely to aggregate funds for a female to stroke care as compared with a male primary wage earner whose disability would lead to substantial loss of household income. Female participants may also lack caregiver support for poststroke recovery.⁶

Because our study did not investigate health care access, female autonomy to make health care decisions, individuals' financial capacity, and cultural perceptions of medical practice in Zambia, we cannot empirically report reasons behind these observed disparities.

This study highlights clinical differences in stroke risk factors and subtypes among male and female individuals and suggests the need to explore complex social and cultural factors that may influence outcomes based on sex differences in stroke care delivery in Zambia.

Appendix. Authors

Study Funding

This work was supported by the American Academy of Neurology Medical Student Research Scholarship, U.S. Department of State Fulbright Scholar Fellowship, and NIH [Grant Numbers: R21 NS118543-01, 1K01TW011771-01A1, and D43TW009340].

Disclosure

A. Nutakki received the American Academy of Neurology Medical Student Scholarship and NIH Fogarty Fellowship [Grant Number D43TW009340]. L. Chishimba reports no disclosures relevant to the manuscript. M. Chomba reports no disclosures relevant to the manuscript. S. Zimba reports no disclosures relevant to the manuscript. R.F Gottesman reports no disclosures relevant to the manuscript. M. N. Bahouth received NIH Grant No. R21 NS118543-01. D. Saylor received the U.S. Department of State Fulbright Scholar Fellowship and NIH Grant No. 1K01TW011771-01A1. Go to Neurology.org/N for full disclosures.