Gender differences in health-seeking behaviour: insights from the National Health and Morbidity Survey 2019

Abstract

Background

Health-seeking behaviour involves actions taken by individuals who feel unwell to seek remedies and varies based on cognitive and non-cognitive factors like sex, age, socioeconomic status, and access to healthcare. Gender roles significantly impact health outcomes with the COVID-19 pandemic further accentuating the gender disparity in public health compliance. Using secondary data from a national health survey, this article aims to assess the gender-based characteristics and factors influencing health-seeking behaviour among the population in Malaysia.

Methods

This study was a secondary data analysis of the NHMS 2019, a cross-sectional national health survey using Andersen’s Behavioural Model. It evaluated factors influencing self-medication and treatment-seeking based on socio-demographics, enabling conditions, and perceived health needs. Descriptive statistics and logistic regression models were conducted to identify factors influencing health-seeking behaviour among men and women.

Results

This study analysed the health-seeking behaviours of 10,933 Malaysian adults, representing 19.7 million people. The overall prevalence of sickness was 16.1% (95% CI = 14.8–17.4), with higher rates in women (18.1%; 95% CI = 95% CI = 16.3–19.9). Among those who were sick, 56.4% (95% CI = 52.9–60.0) sought professional treatment while 23.0% (95% CI = 19.8–26.2) self-medicated. Regression analysis showed that urban women and those rating their health poorly were more likely to seek professional care, while Chinese, those with no formal education, and retiree women were more likely to self-medicate. Among males, those with long-term condition are more likely to seek treatment while students were less likely to self-medicate compared to private employees.

Conclusion

The study reveals significant gender and sociodemographic disparities in health-seeking behaviour amongst Malaysian men and women. The factors that influence these health-seeking behaviour is unique for each gender. This emphasises the importance of targeted interventions which are gender-sensitive to address structural inequities and achieve equitable healthcare utilisation across all demographic groups in Malaysia.

Background

Health or care-seeking behaviour is described as actions taken by individuals who perceive themselves to be ill or have a health condition to find a remedy [1]. The decision made by an individual to engage or utilise healthcare services is heterogeneous and influenced by both intrinsic and extrinsic factors. These factors include sex, age, socioeconomic status, presence of chronic illness or comorbidities, access to service, quality of services provided, and awareness of health status [1, 2]. Gender should be recognised as an important social determinant of health for us to incorporate how gendered factors affect the health outcomes of both men and women [3]. Sex is characterised by biological and genetic traits, while gender is defined as socially constructed roles, norms, behaviours, and attributes that a given society deems appropriate for men and women [4, 5]. While the distinctions between sex and gender are well established, and acknowledging that gender can be non-binary, in the Malaysian context, gender categorisation remains predominantly binary—male or female—and is widely adopted in legal documents and institutional practices. This binary framework influences data collection and reporting, thereby informing the scope of this study.

Gender roles, identity, and relations affect behaviour, access, and use of healthcare systems which accentuates significant disparities in health outcomes for men and women. These gender constructs interact with social norms and power dynamics, influencing how individuals engage with healthcare and their susceptibility to health issues [4]. For example, women are known to be more predisposed to certain autoimmune disorders but also seem to exhibit behaviours that lean towards frequent healthcare utilisation. Men, on the other hand, who tend to suffer from diseases associated with risky lifestyle behaviours such as tobacco smoking and alcohol consumption, often delay care until acute or severe issues arise [6]. The higher risks of morbidity and premature mortality among men for communicable and noncommunicable diseases contribute to a significant disparity in life expectancy between men and women [7]. These global findings are also observed in Malaysia, where men’s life expectancy is approximately 3 years shorter than that of women [8].

Applying a gender lens can help identify factors influencing health burden, access to health services, and response to interventions. Multiple factors act as barriers and facilitators for men and women in utilising healthcare. Literature indicates that women generally demonstrate higher rates of healthcare-seeking behaviour and utilisation compared to men across all health concerns, including mental health services [9–13]. Men’s underutilisation of healthcare services has been linked to societal pressures, including traditional masculinity norms and the stigma associated with seeking help [14, 15]. These pressures often result in men neglecting preventive care and delaying treatment until conditions become severe. Conversely, cultural norms that position women as primary caretakers and their involvement in regular maternal health check-ups contribute to their higher rates of healthcare utilisation [9, 10]. The coronavirus disease 2019 (COVID-19) pandemic highlighted the importance of gender in public health compliance, with studies showing women were more likely to adhere to public health measures due to their risk-averse nature, which underscores the need for gender-sensitive health policies [16].

In Malaysia, gender roles are heavily influenced by traditional and cultural norms. Men are often considered to be household heads, decision makers and breadwinners of the family, while women tend to assume caregiving roles within the family [17]. Research posits that young Malaysian men associate masculine traits to financial independence, physical strength and leadership [18]. These entrenched beliefs prevent men from showing their vulnerability or seeking health advice when they are unwell. Women, on the other hand, are seen to be engaging in health-seeking behaviour especially when there is presence of autonomy and social support [19]. Cultural beliefs, caregiving responsibilities, self-rated health, are factors that influence their health-seeking behaviour. Additionally, it is reported that female-headed households in Malaysia tend to allocate a greater proportion of their income to essential needs such as food and health compared to men who spend on transport and recreation [20].

In this study, we applied Andersen’s Behavioural Model of Health Services Use to guide our understanding of healthcare utilisation. This model suggests that an individual’s decision to seek healthcare is influenced by several factors such as sociodemographic characteristics, enabling, and health needs [9, 21]. For instance, despite the availability of family planning services, economic constraints and authority dynamics linked to gender norms may prevent women from accessing these services [22, 23]. By promoting an equity lens into disease prevention, and optimal management of chronic conditions, insights into healthcare service utilisation can inform effective health policy and intervention strategies towards strengthening equitable healthcare, and recognising gendered determinants of health can help enhance overall healthcare services utilisation.

Healthcare services in Malaysia are provided by both the government and private sector, aiming to achieve universal health coverage by ensuring reasonable access to a large portion of the population [24]. The National Health and Morbidity Survey (NHMS) conducted by the Ministry of Health Malaysia, is the primary source of information on a wide range of health topics including health status, healthcare utilisation, disease prevention, and health-seeking behaviour. Findings from this survey have significantly influenced policy decisions and strategies aimed at improving health systems and population outcomes [25].

While there is ample literature on gender disparities in health-seeking behaviour and utilisation, the largely different healthcare systems between each country and the nature of studies being conducted in various settings limit the ability to identify factors that influence health-seeking behaviours among the Malaysian population. This information is crucial for policymakers and stakeholders in designing strategies that meet community health expectations and outcomes. By applying the principles of gender equity and utilising national survey data from Malaysia, we aim to determine the characteristics of respondents based on gender, assess the prevalence of illness among adults, based on gender, and identify factors influencing health-seeking behaviour among those who reported recent sickness, stratified by gender.

Methods

This study is a secondary data analysis from NHMS 2019, a nationwide cross-sectional population survey among the non-institutionalised general population of Malaysia. The survey employed a random two-stage stratified, proportionate-to-size sampling design to select a national representative sample. The sampling frame was provided by the Department of Statistics of Malaysia, based on the National Population and Housing Census 2010. This study included all states and federal territories in Malaysia, both urban and rural areas, including over 75,000 Enumeration Blocks (EBs). Stratification was done by state (Primary Stratum), then by urban or rural areas within each primary stratum (Second Stratum). Sampling involved a random selection of 463 EBs, within which 14 Living Quarters were chosen. All selected households and eligible respondents in the household were invited to participate.

Data collection took place between 14th July and 2nd October 2019. Trained research assistants carried out face-to-face interviews using mobile tablet devices, guided by a questionnaire system application developed in-house. The structured questionnaire was pre-tested, pilot-tested, and made available in both Malay and English languages [25, 26]. Data collection team leaders arranged appointments with eligible households and made multiple visits, at least three, to ensure good coverage of all eligible respondents. This study focused on respondents with complete data on health-seeking behaviour, recent health problems experienced, and health-seeking behaviour (seeking treatment from healthcare practitioners or self-medicated). A total of 10,933 adults aged 18 years and above were included in this analysis. Detailed methodology and sampling design are described in the NHMS 2019 official report [27].

Study variables

Research on healthcare utilisation from a behavioural perspective aims to identify the most appropriate variables affecting the selection of healthcare services providers [2]. Andersen’s Behavioural Model on Health Services Use encapsulates this by highlighting how predisposition, barriers and facilitators, and care needs predict health-seeking behaviour and health services utilisation [28].

Outcome variables

This study adapted Andersen’s Behavioural Model and looks at health-seeking behaviour which comprises two outcome variables: [1] self-medication among those who reported sickness for the last two weeks before the interview, and [2] seeking treatment from healthcare practitioners among those who reported sickness for the last two weeks before the interview. This was explored through the question “In the last 2 weeks, did you experience any of the following health problems such as fever, sore throat, difficulty in swallowing, running nose or blocked nose, cough, and others” with an option of “Yes” or “No” Those who answered “Yes” were then asked, “In the last 2 weeks, did you seek treatment/medication or advice from healthcare practitioners?” and “In the last 2 weeks, did you take medicine without advice from healthcare practitioners?”. Healthcare practitioners here refer to modern healthcare practitioners including community pharmacists, as well as traditional and complementary medicine practitioners (e.g. spiritual healers, Chinese herbalists, Ayurvedic practitioners, and Islamic medicine practitioners). Self-medication was used to refer to those who answered “Yes” to taking medicine without advice from healthcare practitioners.

Factors influencing health-seeking behaviour

Predisposing factors

The predisposing factors studied were age, ethnicity, education level, marital status, locality, and living conditions. Age was initially measured as a continuous variable and coded into three groups: “18–39”, “40–59” and “60 and Above” based on age distribution pattern. Ethnicity was self-reported identification from one of the following groups: Malay, Chinese, Indian, Aborigines, Bumiputera of Sabah, Bumiputera of Sarawak, and Others; and was then collapsed into “Malay” (i.e. Malay, Aborigines, Bumiputera of Sabah, Bumiputera of Sarawak), “Chinese”, “Indian”, and “Others” to improve distribution. Education was categorised into four groups: “No Formal Education”, “Primary Education”, “Secondary Education”, and “Tertiary Education” Marital status was either “Single”, “Married”, or “Widow(er)/Divorcee” The locality was divided into “Urban” or “Rural” based on the population density of the area. Living conditions were either “Living Alone” or “With Family Members”.

Enabling factors

The enabling factors included employment status, household income, and healthcare coverage. Employment status was classified into “Government Employee”, “Private Employee/Self-employed”, “Not Working”, “Retiree”, and “Student”. Household income was based on total monthly household income, adjusted on an adult-equivalent basis using the parametric equivalence scale that accounts for the composition and size of the household and economies of scale [29]. The household incomes were then grouped into quintiles with the first quintile (Q1) representing the poorest 20% of the population, and the last quintile (Q5), the richest. Subsequently, the quintiles were grouped into three categories: Bottom 40% (those in Q1 and Q2) which refers to the poorest 40%, Middle 40% (those in Q3 and Q4), and Top 20% (those in Q5).

Malaysia adopts a mixed healthcare financing system that comprises public subsidies and private coverage schemes. Public coverage includes government guarantee letters, pensioner cards, and financial subsidies for low-income groups through targeted programmes while private coverage comprises employer-provided medical benefits and private health insurance. In this study, healthcare coverage was defined as having supplementary financial coverage for healthcare which can include a government guarantee letter, pensioner card, government-specific health fund, personal health insurance, employer-provided medical benefits, employer-sponsored health insurance, or a panel clinic.

Health need factors

Perceived and evaluated health needs were used as proxy measures for the need factors. For perceived health needs, self-rated health status was assessed using a five-point scale with the question “How would you rate your health status?” and categorised into three groups: “Good to Excellent”, “Fair”, and “Poor to Very Poor”. Evaluated health need was defined as the presence of at least one long-term condition diagnosed by a doctor or healthcare professional. The questions asked were “Have you ever been told by a doctor or Assistant Medical Officer that you have diabetes?”; “Have you ever been told by a doctor or Assistant Medical Officer that you have high blood pressure?”; and “Have you ever been told by a doctor or Assistant Medical Officer that you have high cholesterol?”, to which the answers were either “Yes” or “No”.

Statistical analysis

Descriptive statistics adjusted for the complex sample design were used to illustrate the sociodemographic characteristics of the respondents, overall and according to gender. We presented the prevalence of those who reported recent sickness and the health-seeking behaviour among those who reported recent sickness, overall and according to gender. The variables selected for inclusion in the regression models were based on statistical criteria and theoretical relevance by adopting the purposeful selection approach as outlined by Bursac et al. [30, 31]. Each independent variable was subjected to univariate logistic regression analysis and variables with a p-value < 0.25 were included in the multivariable model. In the multivariable model, non-significant variables (p > 0.10) were sequentially removed, unless their exclusion impacted coefficient estimates of the other variables by more than 15–20%, indicating confounding. Theoretical relevance by retaining variables that were supported by literature regardless of their statistical significance was adopted as well to ensure a balance between data-driven selection and the existing body of knowledge to produce a parsimonious model [17, 19, 32–35].

Multivariable logistic regression analyses were conducted using the complex survey design to determine the factors influencing health-seeking behaviour, according to gender. The crude odds ratio (COR) and adjusted odds ratio (AOR) with 95% confidence intervals (CI) for the logistic models were displayed. The Variance Inflation Factor (VIF) was used to analyse the multicollinearity. A VIF greater than 10 suggests that multicollinearity may be an issue [36]. Archer-Lemeshow statistics were used to test the goodness of fit model, and a p-value > 0.05 was deemed to be a satisfactory fit [37]. We conducted the analyses using StataCorp Stata 14 (StataCorp, College Station, TX, USA) at a significance level of 5%.

For descriptive analyses, a full case analysis approach was applied by including missing values as a separate category, allowing all respondents to be retained in the analysis. For regression analyses, an available case analysis approach was used, whereby respondents with missing data on any of the model variables were excluded from that specific analysis.

Results

Overall, a total of 10,933 respondents were included in this study, representing a 19.7 million population, consisting of 47.0% and 53.0% men and women respectively. The characteristics of respondents are shown in Table 1.

Table 1.

Characteristics of respondents, stratified by gender (n = 10,933)

Characteristics	All (n = 10,933)		Male (n = 5,138)		Female (n = 5,795)
	Count, n	Percentage (%)	Count, n	Percentage (%)	Count, n	Percentage (%)
Predisposing Factors
Age group (years)
18–39	4542	51.89	2183	52.85	2359	50.99
40–59	3919	31.53	1839	31.63	2080	31.43
60 and above	2472	16.58	1116	15.52	1356	17.57
Ethnicity
Malay	8682	69.03	4083	68.34	4599	69.67
Chinese	1471	24.12	717	25.35	754	22.97
Indian	731	6.47	312	6.05	419	6.87
Others	49	0.37	26	0.26	23	0.48
Education level
No Formal Education	551	3.76	152	2.21	399	5.21
Primary Education	2255	17.23	1007	15.99	1248	18.39
Secondary Education	5312	50.46	2687	54.21	2625	46.94
Tertiary Education	2791	28.31	1282	27.24	1509	29.32
Missing	24	0.24	10	0.35	14	0.13
Marital status
Single	2342	28.68	1324	33.48	1018	24.16
Married	7381	62.73	3572	62.76	3809	62.70
Widow(er)/Divorcee	1203	8.52	238	3.71	965	13.03
Missing	7	0.08	4	0.05	3	0.11
Locality
Urban	6540	75.96	3061	78.34	3479	73.73
Rural	4393	24.04	2077	21.66	2316	26.27
Living condition
Living Alone	1086	11.10	530	12.48	556	9.80
With Family Members	9847	88.90	4608	87.52	5239	90.20
Enabling Factors
Employment status
Government Employee	1203	7.86	572	7.63	631	8.07
Private Employee/Self-employed	5090	52.36	3261	68.53	1829	37.17
Not Working	3774	31.30	768	13.86	3006	47.68
Retiree	544	3.94	398	5.89	146	2.11
Student	320	4.51	137	4.03	183	4.96
Missing	2	0.03	2	0.06	0	0.00
Household income
Bottom 40%	4565	37.40	2095	35.73	2470	38.97
Middle 40%	4358	40.39	2043	39.47	2315	41.26
Top 20%	2010	22.20	1000	24.80	1010	19.77
Healthcare coverage
No	5009	43.98	2150	40.29	2859	47.46
Yes	5789	54.69	2932	58.59	2857	51.02
Missing	135	1.33	56	1.13	79	1.52
Health Need Factors
Self-rated health
Good to Excellent	8141	77.89	3942	79.21	4199	76.65
Fair	2454	19.39	1043	18.09	1411	20.61
Poor to Very Poor	268	2.07	119	2.04	149	2.10
Missing	70	0.65	34	0.66	36	0.64
Presence of at least one long-term condition
No	7508	73.91	3620	74.85	3888	73.03
Yes	3224	24.12	1396	22.72	1828	25.43
Missing	201	1.97	122	2.43	79	1.53

Malay included Aborigines, Bumiputera of Sabah, and Bumiputera of Sarawak

The data (Table 2) shows an overall sickness prevalence of 16.1% (95% CI = 14.8–17.4) among Malaysian adults. However, this rate varies between genders, with 18.1% (95% CI = 16.3–19.9) of women reporting sickness compared to 14.1% (95% CI = 12.6–15.5) of men. Adults aged 60 and above have the highest prevalence of reported sickness at 18.9% (95% CI = 16.4–21.3) and this trend is consistent across both men and women. Women with lower educational levels, particularly those with no formal education (24.1%; 95% CI = 17.6–30.7), report higher sickness prevalence compared to men (22.0%; 95% CI = 13.6–30.4). Additionally, rural residents report slightly higher sickness prevalence (17.5%; 95% CI = 15.4–19.7) compared to urban residents (15.7%; 95% CI = 14.1–17.3), with more significant gender-based disparities in rural settings. The data also demonstrates that almost half of the individuals with poor self-rated health were more likely to report being sick (48.2%; 95% CI = 39.5–57.0) with women showing a slightly higher likelihood (50.0%; 95% CI = 37.6–62.5) of reporting sickness than men (46.6; 95% CI = 35.1–58.1).

Table 2.

Prevalence of those who reported recent sickness, stratified by gender (n = 10,933)

Characteristics	Overall (n = 10,933)		Male (n = 5,138)		Female (n = 5,795)
	Count, n (Unweighted)	% Weighted (95% CI)	Count, n (Unweighted)	% Weighted (95% CI)	Count, n (Unweighted)	% Weighted (95% CI)
OVERALL	2,032	16.1 (14.8–17.4)	818	14.1 (12.6–15.5)	1,214	18.1 (16.3–19.9)
Predisposing Factors
Age group (years)
18–39	839	15.7 (14.0–17.5)	331	13.3 (11.1–15.5)	508	18.1 (15.6–20.6)
40–59	681	15.4 (13.6–17.1)	267	13.3 (11.1–15.6)	414	17.3 (15.0–19.6)
60 and above	512	18.9 (16.4–21.3)	220	18.2 (14.8–21.6)	292	19.4 (16.0–22.9)
Ethnicity
Malay	1,723	17.9 (16.4–19.4)	683	15.0 (13.4–16.6)	1,040	20.5 (18.5–22.5)
Chinese	164	10.8 (8.3–13.3)	83	11.8 (8.1–15.5)	81	9.7 (6.6–12.9)
Indian	139	17.3 (12.8–21.8)	48	12.3 (7.8–16.8)	91	21.5 (15.7–27.3)
Others	6	22.3 (1.4–43.1)	4	22.4 (8.3–36.5)	2	22.2 (0.0–52.4)
Education level
No Formal Education	129	23.5 (18.3–28.7)	32	22.0 (13.6–30.4)	97	24.1 (17.6–30.7)
Primary Education	451	19.3 (16.8–21.7)	180	18.5 (15.0–21.9)	271	19.9 (16.9–22.9)
Secondary Education	913	14.6 (13.1–16.2)	399	12.8 (11.0–14.5)	514	16.6 (14.4–18.8)
Tertiary Education	532	16.0 (13.6–18.3)	204	13.3 (10.1–16.4)	328	18.3 (15.0–21.6)
Missing	7	19.1 (0.0–39.4)	3	18.3 (0.0–44.7)	4	21.2 (0.0–44.2)
Marital status
Single	382	13.7 (11.5–15.9)	172	11.0 (8.9–13.2)	210	17.2 (13.5–20.9)
Married	1,372	16.6 (15.1–18.0)	595	15.1 (13.3–17.0)	777	17.9 (16.0–19.8)
Widow(er)/Divorcee	276	20.8 (17.1–24.5)	50	22.1 (14.4–29.8)	226	20.5 (16.4–24.6)
Missing	2	50.0 (3.0–97.1)	1	78.3 (37.7–100)	1	38.4 (0.0–96.0)
Locality
Urban	1,242	15.7 (14.1–17.3)	485	13.3 (11.7–15.0)	757	18.0 (15.9–20.2)
Rural	790	17.5 (15.4–19.7)	333	16.6 (13.9–19.4)	457	18.2 (15.3–21.1)
Living condition
Living Alone	199	15.9 (12.3–19.4)	83	15.1 (9.7–20.5)	116	16.8 (12.5–21.1)
With Family Members	1,833	16.2 (14.8–17.6)	735	13.9 (12.4–15.4)	1,098	18.2 (16.4–20.1)
Enabling Factors
Employment status
Government Employee	272	17.9 (14.4–21.5)	99	12.9 (9.1–16.8)	173	22.4 (16.6–28.1)
Private Employee/Self-employed	860	15.5 (13.7–17.2)	488	14.1 (12.2–16.0)	372	17.9 (15.0–20.8)
Not Working	756	17.5 (15.6–19.4)	145	15.1 (11.9–18.2)	611	18.2 (16.0–20.4)
Retiree	93	14.7 (10.5–18.9)	66	13.7 (8.4–19.0)	27	17.2 (10.1–24.4)
Student	51	12.4 (7.6–17.1)	20	12.9 (5.6–20.2)	31	12.0 (6.0–17.9)
Household income
Bottom 40%	811	16.2 (14.4–18.1)	309	13.1 (10.9–15.3)	502	19.0 (16.6–21.4)
Middle 40%	825	16.0 (14.2–17.7)	333	13.6 (11.8–15.4)	492	18.1 (15.5–20.7)
Top 20%	396	16.2 (13.5–19.0)	176	16.2 (12.6–19.7)	220	16.3 (12.7–20.0)
Healthcare coverage
No	892	15.7 (14.2–17.2)	334	14.1 (12.0–16.1)	558	17.1 (15.1–19.1)
Yes	1,117	16.4 (14.7–18.2)	475	14.0 (12.0–16.0)	642	19.1 (16.4–21.7)
Missing	23	18.0 (9.2–26.8)	9	19.0 (4.5–33.5)	14	17.4 (6.9–27.8)
Health Need Factors
Self-rated health
Good to Excellent	1,142	12.3 (11.1–13.6)	485	11.0 (9.4–12.5)	657	13.6 (11.9–15.4)
Fair	756	27.9 (25.2–30.6)	274	23.6 (20.2–27.1)	482	31.4 (27.8–35.0)
Poor to Very Poor	121	48.2 (39.5–57.0)	53	50.0 (37.6–62.5)	68	46.6 (35.1–58.1)
Missing	13	19.6 (5.8–33.5)	6	9.3 (0.0–18.8)	7	29.6 (6.3–52.9)
Presence of at least one long-term condition
No	1,253	14.6 (13.2–16.0)	495	12.2 (10.5–13.8)	758	16.9 (15.0–18.8)
Yes	748	21.5 (19.1–23.9)	308	21.2 (17.6–24.7)	440	21.8 (18.6–24.9)
Missing	31	9.9 (5.3–14.5)	15	5.8 (1.3–10.3)	16	16.0 (6.8–25.2)

Malay included Aborigines, Bumiputera of Sabah, and Bumiputera of Sarawak

% Percentage, CI Confidence Interval

Among those who reported sickness, in general, about 56.4% (95% CI = 52.9–60.0) reported seeking treatment from healthcare practitioners while 23.0% (95% CI = 19.8–26.2) self-medicated for their sickness (Table 3). While there were some differences between men and women in their health-seeking behaviour, the overall patterns were similar.

Table 3.

Health-seeking behaviour among those who reported recent sickness, stratified by gender (n = 2,032)

Health-seeking behaviour	Overall (n = 2,032)		Male (n = 818)		Female (n = 1,214)
	Count,n (Unweighted)	% Weighted (95% CI)	Count,n (Unweighted)	% Weighted (95% CI)	Count,n (Unweighted)	% Weighted (95% CI)
Sought treatment from healthcare practitioner
Yes	1,156	56.4 (52.9-60.0)	442	53.4 (48.1-58.8)	714	58.6 (54.1-63.2)
No	868	42.8 (39.3-46.4)	372	45.9 (40.5-51.2)	496	40.6 (36.1-45.2)
Missing	8	0.7 (0.0-1.4)	4	0.7 (0.0-1.5)	4	0.7 (0.0-1.7)
Self-medicated
Yes	457	23.0 (19.8-26.2)	187	22.2 (17.6-26.9)	270	23.5 (19.6-27.5)
No	1,564	76.1 (72.9-79.3)	626	77.0 (72.4-81.6)	938	75.4 (71.4-79.5)
Missing	11	0.9 (0.1-1.7)	5	0.7 (0.0-1.6)	6	1.0 (0.0-2.3)

% Percentage, CI Confidence Interval

Table 4 displays multivariable logistic regression results of health-seeking behaviours among male and female adults in Malaysia. Model I and II assessed the factors influencing male and female adults seeking treatment from healthcare practitioners, respectively. Among males, those with at least one long-term condition (AOR = 1.67; 95% CI = 1.03–2.69) are more likely to seek treatment from healthcare practitioners as compared to those without. Among females, urban dwellers (AOR = 1.57; 95% CI = 1.06–2.33) are more likely to seek treatment from healthcare practitioners when compared to rural dwellers. Females who self-rated their health as poor to very poor were 4 times more likely to seek treatment from healthcare practitioners than those who rated their health as good to excellent (AOR = 3.84, 95%CI = 1.80–8.18).

Table 4.

Logistic regression model for health-seeking behaviours among those who reported recent sickness, stratified by gender

Characteristics	Sought treatment from healthcare practitioner				Self-medicated
	Model I-Male		Model II-Female		Model III-Male		Model IV-Female
	COR	AOR	COR	AOR	COR	AOR	COR	AOR
	(95% CI)	(95% CI)	(95% CI)	(95% CI)	(95% CI)	(95% CI)	(95% CI)	(95% CI)
Age group (years)
18–39	1.00 (ref)		1.00 (ref)		1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)
40–59	0.87 (0.54–1.40)		1.09 (0.77–1.56)		1.05 (0.60–1.84)	1.05 (0.59–1.89)	1.17 (0.76–1.80)	1.22 (0.75–2.01)
60 and above	1.86 (1.12–3.12)*		0.87 (0.52–1.45)		0.82 (0.44–1.53)	0.98 (0.47–2.02)	1.08 (0.61–1.89)	0.56 (0.27–1.15)
Ethnicity
Malay	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)
Chinese	0.78 (0.39–1.54)	0.68 (0.35–1.34)	0.87 (0.42–1.82)	0.81 (0.38–1.72)	1.26 (0.56–2.84)	1.13 (0.54–2.35)	1.78 (0.88–3.59)	2.75 (1.39–5.47)**
Indian	1.56 (0.67–3.63)	1.61 (0.61–4.26)	1.30 (0.73–2.31)	1.20 (0.67–2.12)	0.86 (0.31–2.37)	0.84 (0.29–2.46)	0.73 (0.38–1.38)	0.81 (0.42–1.55)
Others	1.73 (0.17–17.44)	1.16 (0.11–12.64)	1.00 (Not estimable)	1.00 (Not estimable)	1.67 (0.17–16.86)	1.41 (0.15–13.28)	4.38 (0.26–72.65)	8.02 (0.16–405.53)
Education level
No Formal Education	0.64 (0.22–1.81)		0.82 (0.42–1.63)		0.64 (0.21–2.00)	0.86 (0.26–2.81)	1.88 (0.92–3.86)	3.18 (1.39–7.23)**
Primary Education	1.09 (0.60–1.99)		0.80 (0.48–1.33)		0.71 (0.34–1.50)	0.80 (0.35–1.82)	1.41 (0.76–2.62)	1.91 (0.93–3.91)
Secondary Education	0.98 (0.59–1.62)		0.64 (0.40–1.02)		0.73 (0.39–1.38)	0.84 (0.44–1.59)	0.88 (0.50–1.52)	0.90 (0.50–1.62)
Tertiary Education	1.00 (ref)		1.00 (ref)		1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)
Marital status
Single	1.00 (ref)		1.00 (ref)		1.00 (ref)		1.00 (ref)	1.00 (ref)
Married	0.72 (0.45–1.15)		1.25 (0.76–2.06)		1.19 (0.68–2.06)		0.61 (0.34–1.08)	0.54 (0.28–1.03)
Widow(er)/Divorcee	0.88 (0.40–1.95)		1.16 (0.60–2.22)		0.53 (0.20–1.43)		0.83 (0.41–1.69)	0.63 (0.28–1.43)
Locality
Urban	1.08 (0.69–1.68)		1.41 (0.98–2.04)	1.57 (1.06–2.33)*	0.83 (0.52–1.31)		1.11 (0.72–1.71)
Rural	1.00 (ref)		1.00 (ref)	1.00 (ref)	1.00 (ref)		1.00 (ref)
Living condition
Living Alone	1.00 (ref)		1.00 (ref)		1.00 (ref)		1.00 (ref)
With Family Members	1.00 (0.47–2.14)		0.57 (0.32–1.01)		0.66 (0.25–1.80)		1.13 (0.62–2.08)
Employment status
Government Employee	1.08 (0.51–2.28)	0.89 (0.42–1.91)	1.23 (0.72–2.10)	1.35 (0.79–2.29)	1.05 (0.43–2.55)	1.04 (0.40–2.69)	0.91 (0.45–1.87)	1.28 (0.62–2.64)
Private Employee/Self-employed	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)
Not Working	2.16 (1.30–3.58)**	1.50 (0.86–2.63)	0.99 (0.67–1.48)	1.01 (0.67–1.52)	0.70 (0.38–1.29)	0.78 (0.42–1.46)	1.00 (0.60–1.67)	1.04 (0.63–1.74)
Retiree	1.76 (0.63–4.95)	1.37 (0.49–3.80)	1.05 (0.41–2.71)	0.75 (0.27–2.06)	0.72 (0.29–1.79)	0.68 (0.24–1.91)	3.69 (1.30–10.45)*	6.88 (2.22–21.34)**
Student	1.82 (0.52–6.40)	2.27 (0.60–8.63)	0.68 (0.25–1.86)	0.67 (0.24–1.89)	0.02 (0.01–0.10)**	0.02 (0.01–0.10)**	1.01 (0.34–2.99)	0.97 (0.31–2.97)
Household income
Bottom 40%	0.89 (0.49–1.63)		0.88 (0.49–1.60)		0.67 (0.32–1.38)	0.74 (0.38–1.44)	1.48 (0.75–2.94)	1.82 (0.90–3.68)
Middle 40%	0.86 (0.47–1.56)		1.09 (0.60–1.98)		0.57 (0.27–1.20)	0.61 (0.31–1.19)	1.64 (0.79–3.38)	1.92 (0.95–3.92)
Top 20%	1.00 (ref)		1.00 (ref)		1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)
Healthcare coverage
Yes	1.01 (0.67–1.52)		0.93 (0.66–1.33)		0.90 (0.53–1.53)		1.25 (0.81–1.91)
No	1.00 (ref)		1.00 (ref)		1.00 (ref)		1.00 (ref)
Self-rated health
Good to Excellent	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)		1.00 (ref)
Fair	1.40 (0.91–2.14)	1.14 (0.73–1.77)	1.24 (0.88–1.74)	1.26 (0.88–1.79)	0.86 (0.51–1.43)		1.02 (0.68–1.53)
Poor to Very Poor	3.04 (1.42–6.51)**	2.17 (0.96–4.89)	3.93 (1.86–8.30)**	3.84 (1.80–8.18)**	0.86 (0.35–2.14)		0.59 (0.28–1.22)
Presence of at least one long-term condition
Yes	1.80 (1.15–2.84)*	1.67 (1.03–2.69)*	1.37 (0.90–2.08)		0.70 (0.42–1.18)		1.03 (0.68–1.56)
No	1.00 (ref)	1.00 (ref)	1.00 (ref)		1.00 (ref)		1.00 (ref)

Malay included Aborigines, Bumiputera of Sabah, and Bumiputera of Sarawak

Archer-Lemeshow: Model I (p = 0.124), Model II (p = 0.811), Model III (p = 0.122), Model IV (p = 0.128) (representing an excellent fit). Multicollinearity was unlikely (VIF < 5.0)

Results for ‘Others’ could not be estimated due to the small sample size

COR Crude Odd Ratio, AOR Adjusted Odd Ratio, CI Confidence Interval, ref reference

*p-value < 0.05; **p-value < 0.001

Model III and IV assessed the factors influencing self-medication behaviours among male and female adults, respectively. Among males, students (AOR = 0.02; 95% CI = 0.01–0.10) were less likely to self-medicate than private employees. Among females, Chinese were shown to be 3 times (AOR = 2.75; 95%CI = 1.39–5.47) more likely to self-medicate than Malays. Females with no formal education were also 3 times (AOR = 3.18; 95% CI = 1.39–7.23) more likely to self-medicate as compared to those with tertiary education. Additionally, retiree females are 7 times (AOR = 6.88; 95% CI = 2.22–21.34) more likely to self-medicate as compared to private employees.

Discussion

This study offers comprehensive insights into gender differences in health-seeking behaviour in Malaysia, drawing from NHMS 2019 data. Our findings depict women reported a higher prevalence of sickness (18.1%) compared to men (14.1%), consistent with global trends where women typically report more health issues and utilise healthcare services more frequently [9, 10, 12, 13, 38]. Additionally, sickness prevalence was documented to be higher among women not working, with no formal education, and of lower socio-economic status compared to men. This disparity in employment status and education level between both genders, further explained by the differences in income status amongst women compared to men, can contribute to poorer health literacy and financial freedom that ultimately results in poorer health outcomes [39, 40].

Notably, this study found a greater percentage of healthcare utilisation among the local women population compared to men, aligning with the global trend of healthcare utilisation among women [13, 41, 42]. This corroborates recent findings on health-seeking behaviour among women in urban poor women in Malaysia as well as 72.4% of these women sought care when unwell [19]. Furthermore, age-related health concerns are likely driving this behaviour as older women who usually experience more comorbid conditions, demonstrated higher healthcare-seeking behaviour compared to men. Previous NHMS findings from 2006 indicated that the Chinese population in Malaysia predominantly sought treatment from private healthcare facilities [43]. In contrast, our study revealed that Chinese women are now more likely to self-medicate. This shift highlights that health-seeking behaviour is dynamic and influenced by evolving cultural, social, and systemic factors.

Our findings also indicated that there are higher rates of women who are retired or less educated self-medicating suggesting potential barriers that are present in accessing healthcare services. While factors such as reproductive health needs, higher rates of chronic conditions, and greater health awareness among women influence women’s healthcare utilisation, this behaviour should not be misconstrued as an indication that women have equitable access to healthcare [11, 22, 39, 42, 44–46]. The intersection of geography and gender-based disparities is highlighted in this study as rural women are documented to be less likely to seek healthcare services. Rural provinces, often lack the healthcare infrastructure that caters to the specific healthcare needs of women which can leave women at a particular disadvantage.

Several characteristics among women indicated that structural barriers could be complicating their access to services. Studies globally have indicated that positive health outcomes are associated with the ability of women to make health decisions and household finances [39, 40, 47, 48]. Locally, NHMS 2015 data depicted that only 37.8% of women reported deciding on their own to seek care and almost 18% of healthcare decisions were made by their spouse. This is further compounded by locality factors as 43.5% of individuals decide for themselves in urban settings when to seek care while only 34.7% of those in rural areas do the same [49]. Additionally, local findings report that only 19.6% of households are headed by females and these households had lower average monthly expenditures compared to men highlighting economic disparities that might constrain access to healthcare [20]. The complex interplay of geography, decision-making, economic disadvantages, and the lack of social support, demonstrate the effect of systemic gender inequities present that influence health-seeking behaviour [48, 50, 51].

Through our study, while specific factors drive women’s health-seeking behaviour, these same factors do not appear to significantly impact men suggesting a potential gender difference in the determinants of healthcare utilisation. These findings contrast with previous claims that no gender differences are present in healthcare utilisation and further strengthen the need to embed gender-specific drivers in policies to ensure equitable healthcare utilization [52, 53]. Men may generally delay or avoid seeking care, however presence of chronic illnesses plausibly because of aging increases the likelihood that they will seek treatment due to the need for continuous management and the potential severity of untreated conditions [54]. According to the Labor Force Statistics 2023 in Malaysia, men are more likely to be employed and work longer hours compared to women [55]. Employment-related pressures such as limited access to paid medical leave, fear of income loss, inflexible working hours, and workplace norms that discourage perceived vulnerability may deter men from seeking well when unwell. In our study, we also observed that male students were less likely to self-medicate than those employed in the private sector. Multiple factors such as health literacy levels, cultural norms, parental roles as caregivers, and access to institutional healthcare services could have plausibly influenced this health-seeking behaviour among students [56–59]. Additionally, this pattern may also reflect the early internalisation of masculine norms prevalent among young Malaysian men, potentially making them more cautious about self-medicating while remaining hesitant to engage with formal healthcare services.

Policymakers must be aware of these nuances to ensure a healthcare system that caters to both genders are built by addressing the unique interactions that are present. Targeted interventions such as accessible healthcare facilities, culturally sensitive healthcare services, and strengthened social support systems can promote an equitable health system for both men and women. For example, the National Men’s Health Plan of Action for Malaysia has identified strategies such as health campaigns, promotions, and educational material that contains messages curated specifically for men [60]. By incorporating locality-specific health campaigns to reach rural populations, providing targeted resources for men with lower health literacy, and strengthening community-based support networks, men can be empowered in more positive health-seeking behaviours. The Malaysian government has set aside initiatives that promote gender equity by providing incentives and assistance to women returning to the workforce [61]. While these efforts are commendable, ongoing support in terms of flexible work policies, childcare options, and health benefits should be made accessible to women to foster an inclusive and supportive working environment. To fully realise the benefits of initiatives planned by the government, gender-sensitive policies must be established to ensure these strategies reap their intended outcome.

The dataset for this study was from a national survey that provides findings that are reflective of the broader population, The comprehensive data coverage involving 10,933 adults allows for in-depth subgroup analysis of the factors that influence health behaviours and allows us to identify the complex interaction that is present between gender, socio-economic status, and healthcare preferences. However, due to the nature of the survey being a cross-sectional self-reported questionnaire, recall bias and causal inferences from the findings cannot be made. It is imperative to note that by merely disaggregating data by sex, the complexities of gender power relations and their impact on health are difficult to capture [39]. Gender analysis, by combining both qualitative and quantitative methods in research, is necessary to reveal the underlying social inequities that determine healthcare access and other forms of behaviour [62]. Furthermore, the binary classification of variables such as healthcare coverage could oversimplify the complexity of the healthcare financing landscape in Malaysia and future studies should consider a more detailed categorisation to better understand the role or impact of healthcare coverage on population health behaviours.

Conclusion

This study provides valuable insights into gender differences in health-seeking behaviour amongst Malaysian adults. Women were depicted to be more likely to report sickness and utilise healthcare services compared to men. Through this study, we also observed gender inequities that affect health behaviours for those less educated, rural folks, those not in formal employment, and poorer health. Targeted interventions and gender-sensitive policies to address these inequities are relevant to ensure those in underserved communities are provided appropriate access to healthcare services. Policymakers should adopt a holistic approach that addresses the gender-specific drivers that influence health-seeking behaviour among men and women and must be advocated to meet the needs of all demographic groups.

Abbreviations

COVID-19

Coronavirus Disease 2019

EB

Enumeration Blocks

NHMS

National Health and Morbidity Survey

Authors’ contributions

DSRM and SJ contributed to the conception and study design, analysed and interpreted the data, and prepared the manuscript. All authors contributed to reviewing, editing, read and approved the final version of the manuscript to be published.

Funding

This project was registered under the National Medical Research Registry (NMRR-18-3085-44207) and received funding from the National Institutes of Health, Ministry of Health Malaysia research grant [KKM/NIHSEC/800-3/2/1 Jld.5 [48]. The funding sources were not involved in study design, study conduct, analysis of the results, writing of this manuscript, and the decision to submit this topic for publication.

Data availability

The dataset that supports the findings of this article is not publicly available to protect participant privacy. Request for data can be obtained from the Head of Centre for Biostatistics & Data Repository, National Institutes of Health, Ministry of Health Malaysia on reasonable request and with permission from the Director General of Health, Malaysia.

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki and approved by the Medical Research and Ethics Committee (MREC), Ministry of Health Malaysia [KKM/NIHSEC/P18-2325 [11] dated 20 December 2018, and was registered in the National Medical Research Register, Ministry of Health Malaysia (NMRR-18-3085-44207). This secondary analysis does not require additional ethical approval and administrative permission because the data were anonymised and the authors who were involved in the primary study, hold ownership of the data.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.