School absenteeism during menstruation among adolescent girls in Southern Uganda: A multi-level risk factors analysis

Abstract

Menstrual hygiene management (MHM) continues to be a major barrier impacting adolescent girls’ school attendance in low-income countries. In Uganda, only 10 % of adolescent girls practice adequate MHM; and they miss an estimated 11 % of school days due to menstruation. Missing school (school absenteeism) has been associated with negative educational and social outcomes. This paper aimed to explore the individual, family, school and community-level factors associated with school absenteeism among adolescent girls during menstruation. We used crosssectional (baseline) data from 1,237 adolescent girls recruited from 47 secondary-schools in Southern Uganda. The nested logistic model analysis was used to test the significance of blocks of predictor factors on school absenteeism. The results showed that all blocks of factors are statistically significant. Discrete analysis indicated that dysmenorrhea was associated with an 8.1 % increase in school absenteeism. However, family sufficiency and support appeared to be protective factors decreasing absenteeism by 13.8 % and 4.9 %, respectively. Also, both schools’ good water, sanitation, and hygiene (WASH) infrastructure and supportive teachers were associated with decreasing school absenteeism by 28.6 % and 4.6 %, respectively. Inaddition, cultural taboos on menstruation were associated with the strongest (34.3 %) increase in school absenteeism. Our results showed that all blocks of factors matter when discussing MHM and school absenteeism among adolescent girls. Therefore, more multifaceted interventions are needed to address the issue affecting millions of girls regularly.

1. Background

More than 50 % of adolescent girls in low- and middle-income countries (LMICs) report inadequate Menstrual Hygiene Management (MHM) – defined as “using a clean menstrual management material to absorb or collect menstrual blood, which can be changed in private as needed, using soap and water to wash their bodies as needed, and having access to safe and convenient disposal facilities for used menstrual management materials“ (UNICEF, 2020, p. 17). The majority of those girls live in rural areas (Miiro et al., 2018; Montgomery et al., 2012; Nabwera et al., 2021). Inadequate MHM has been associated with school absenteeism, which in turn, has been shown to negatively impact school-related outcomes, including school performance and grade progression, and increases the odds of dropping out among adolescent girls. (Balkis et al., 2016; Hennegan et al., 2016; Khandaker et al., 2022; Kuhlmann et al., 2017). In Uganda, only 10 % of adolescent girls have been reported to practice adequate MHM, and they miss an estimated 11 % of school days due to menstruation; and that the girls in peri-urban neighborhoods in Uganda missed school on 28 % of period-days, compared with 7 % of non-period days (Miiro et al., 2018). Studies point to risk factors that can be grouped into individual, family, school/structural, and cultural factors contributing to school absenteeism during menstruation in LMICs (Maulingin-Gumbaketi et al., 2022; Sumpter & Torondel, 2013). Among individual factors, first of all, dysmenorrhea and other menstrual-related physical problems such as headaches, stomach and back discomfort, and the heavy flow keep many adolescent girls from attending school during their period (Crofts & Fisher, 2012; De Sanctis et al., 2015; Maqbool et al., 2022; Miiro et al., 2018; Moy & Gupta, 2022; Peuranpää et al., 2014; Tamiru et al., 2015; UNICEF, 2015; Maqbool et al., 2022; Ministry of Health of Uganda, 2017). Furthermore, puberty-specific stressors, menstrual symptoms, and menstrual deficits were associated with depression and anxiety in adolescents girls (Cherenack & Sikkema, 2022), and menstruation-associated stigma and lack of education force many adolescents to miss school (Benshaul-Tolonen et al., 2020; Betsu et al., 2023; Cherenack et al., 2020; Cherenack & Sikkema, 2022; Chinyama et al., 2019; Crichton et al., 2013; Stoilova et al., 2022). In addition, financial constraints related to hygiene products can cause girls to miss school (Dolan et al., 2014; Greed, 2014; Hennegan et al., 2016; Mason et al., 2013).

In the absence of adequate menstrual hygiene products, many girls are forced to improvise with unhygienic or inadequate materials or impose certain restrictions that may negatively impact their health (UNICEF, 2016). However, even though sometimes girls reported using sanitary disposal pads during their last period, only a few of them reported adequate MHM using all components of the standard definition – which implies using only manufactured products (which may be by locally-made re-usable pads), disposing of them in a bin or incinerator, always having access to water and soap at school, and not feeling anxious about their following period (Miiro et al., 2018). Among structural factors, the lack of private washroom spaces with water and soap at schools, and bins or incinerators for disposal have been named as critical reasons for absenteeism among adolescent girls during menstruation (Dolan et al., 2014; Greed, 2014; Sommer, 2013; Sumpter & Torondel, 2013).

Other common risk factors for adolescent girls leaving school include fear of ridicule by classmates (Benshaul-Tolonen et al., 2020; Maqbool et al., 2022; Mason et al., 2013; Ministry of Health of Uganda, 2017; Stoilova et al., 2022). Menstrual stigma and taboo are important sociocultural elements that have a detrimental impact on successful MHM and subsequent school attendance (Maulingin-Gumbaketi et al., 2022). Moreover, teachers’ and families’ reluctance to address menstruation with girls plays a role in perpetuating myths and misconceptions regarding puberty and menstruation (Chandra-Mouli & Patel, 2017).

Given the importance of the aforementioned contributing factors to school absenteeism due to menstrual cycle periods, still only a few studies have explored the comorbidities of various risk factors for school absenteeism among adolescent girls (Ssesanga et al., 2024; Stoilova et al., 2022; Swe et al., 2022). Using the ecological paradigm, this paper contributes to the growing literature and establishes the primary factors at the individual, family, and community levels that contribute to school absenteeism during menstrual hygiene in LIMCs. It could be that all these factors contribute equally, but it could also be that some factors are more pronounced and need faster intervention than others. It could also be that all the observable predictors of school absenteeism, and as a result, menstrual hygiene needs to be addressed concurrently. This question is important given that these are poor communities that must prioritize their use of limited resources.

1.1. Study context

Uganda is a poor, developing country that has made strides in education for the girl child, especially in the past two decades, beginning with the passing of Universal Primary Education (UPE) in 1997, and then Universal Secondary Education (USE) in 2007. Since the introduction of the USE policy, girls are about 49 % more likely to be enrolled in public secondary schools (Asankha & Takashi, 2011). However, the increase in educational attainment remains low (Huylebroeck & Titeca, 2015). Given that adolescent girls miss an estimated 11 % of school days due to menstruation (Hennegan et al., 2016), it is unclear which specific and superior drivers among the four blocks of risk factors (Individual, Family, School, and Community) significantly affect adolescent girls’ school absenteeism. These questions are critical for developing and implementing multifaceted and multilevel strategies to promote adolescent girls’ school attendance, participation, retention, and consequently educational success. To the best of our knowledge, no empirical research in SSA has employed an ecological approach that focuses on various levels of drivers at the same time to fully explain the factors of school absence during menstruation.

1.2. Theoretical framework

To understand the driving factors that impact school attendance during menstruation in the context of rural Uganda, we propose to apply an ecological paradigm and integrate the ecological systems theory of human development (Bronfenbrenner, 1979) with the integrated model of menstruation experiences (Hennegan et al., 2016). The ecological systems model states that determinants of an outcome operate at multiple and interacting levels (Bronfenbrenner, 1979). Based on the qualitative studies in LMICs, the integrated model of menstrual experience describes the potential directional and bidirectional relationships on how the socio-cultural contexts and limited resources affect menstruation experiences that affect girls’ physical and psychological health, social participation, and education-employment (Hennegan et al., 2019).

Based on an integrated model of menstruation experiences and the ecological systems theory, we examine the individual, family, school, and community level factors related to menstruation experiences on school attendance among adolescent girls (Fig. 1). Guided by the ecological model, the blocks of risk factors included are: a) micro system/individual-level factors such as mental health conditions during menstruation, general physical health, dysmenorrhea (pain with menstruation), knowledge of menstruation and menstrual practices; b) meso system/family-level factors such as family cohesion, family support, family communication, and family sufficiency indicators; and c) exosystem /school-level factors and resources such as WASH facilities in school, classmates and teachers support, and gender norms about schooling and education; and d) macro/community level- factors including socio-cultural context/menstruation taboos and community resources/poverty indicators. We intend to test blocks of risk factors (individual, family, school, and community) and discrete factors that contribute to adolescent girls’ school absenteeism during the menstruation period (Bronfenbrenner, 1979).

Fig. 1.

Adolescent girls’ menstruation experiences and school absenteeism in Southern Uganda.

1.3. Study objective

Guided by the ecological framework (Bronfenbrenner, 1979) and integrated models (Hennegan et al., 2019), we tried to answer the following research questions:

Question 1. Which blocks of risk factors (Individual, Family, School, and Community) significantly contributed to adolescent girls’ school absenteeism?

Question 2. What are the individual factors within the different blocks of risk factors that are driving forces for school absenteeism among adolescent girls in Southern Uganda?

2. Materials and methods

2.1. Study sample

We used baseline data collected between July 2018 to February 2019 from NIH funded study called Suubi4Her, funded by the National Institute of Mental Health (NIMH, Grant #: R01MH113486, PI: Fred M. Ssewamala, Ph.D.). The study was conducted in Uganda’s Masaka region, a relatively poor region heavily impacted by poverty and HIV. A total of 1,260 adolescent girls (aged 14–17 at enrollment) were enrolled in the Suubi4Her study. School selection was informed by stratified sampling based on geographic location (rural vs. urban) and student population size (medium vs. large), ensuring balanced representation across different school contexts (Ssewamala et al., 2018).

2.2. Participant recruitment

The following inclusion criteria were used for participants’ inclusion into the study: 1) enrolled in the first or second year of secondary school in one of the 47 secondary schools in the study; 2) attending school in one of the five districts included in the study: Rakai, Kyotera, Masaka, Lwengo or Kalungu; 3) aged 14–17 years; 4) living within a family (not an orphanage/institution, as those in those settings have different familial needs).

All the 47 secondary schools in the study were matched based on socioeconomic status of the students attending these schools, school size (total number of students enrolled), location (urban vs. rural), and overall performance based on the past three years of Uganda Certificate of Education (UCE) examinations, administered by the Uganda Government’s Ministry of Education and Sports. School administrators helped to identify potential participants and their parents/caregivers. Parents/caregivers were given flyers notifying them of the study and inviting them to contact the school head teacher for further details. Caregivers and adolescents who expressed interest were later invited to meet with the in-country project coordinator for a one-on-one informational meeting during which they (parents/caregivers and adolescents) were informed verbally and in writing, the purpose of the study, voluntary participation, extent of their participation, risk and benefits, as well as protection and confidentiality issues. Interested caregivers/parents signed the informed consent and adolescent girls signed the assent forms. Detailed information on participants’ recruitment and selection process are described in the study protocol (Ssewamala, et al., 2018).

2.3. Data collection

Baseline data were collected using a multidimensional survey instrument, which combines existing evidence-based measurement tools, as well as adapted scales and questions explicitly developed for girls in the sub-Saharan Africa (SSA) (Ismayilova et al., 2012; Ssewamala et al., 2010, 2018; Ssewamala & Ismayilova, 2009). All the interviews were conducted in Luganda, the commonly spoken language in the study region. Trained research assistants verbally administered the survey during face-to-face interviews using Qualtrics software on electronic tablets. To ensure consistency and accuracy, the survey instruments were back-translated into English and supervised by trained experts from Makerere University School of Languages, Literature, and Communication in Uganda. The translators held undergraduate degrees specializing in both English and Luganda and were fluent in both languages, having studied English from primary through tertiary levels. All translations were reviewed and cross-checked by at least three team members to ensure consistency of meaning and clarity of expression. While back-translation is a widely accepted strategy for preserving conceptual equivalence, it may not fully capture subtle cultural or contextual nuances, which represents a potential limitation in the linguistic validity of some items.

2.4. Ethics and consent

The study received IRB approval from Washington University Institutional Review Board (IRB-#201703102), the Uganda Virus Research Institute (GC/127/17/07/619), and the Uganda National Council of Science and Technology (SS4406). The study is registered in the Clinical Trials database NCT03307226. Voluntary written informed assent and consent were obtained from the adolescent participants and their caregivers, respectively. Guardians and adolescents provided consent separately to avoid coercion. The research assistants completed training courses on Good Clinical Practice (GCP) and research ethics from the Collaborative Institutional Training Initiative (CITI Program).

2.5. Measures

School Absenteeism assessment:

The outcome variable, missing school due to menstruation, was assessed via two questions: a) In a month, how many days do you miss school because of your menstruation period? (0–10 + days), and b) how often does your menstruation period make you miss school? (Never-1, Once or twice, Several times, Many times, Always-5). A dichotomized variable (no/yes) for school absenteeism was created by merging responses from the two questions: No (coded 0) – if the responses for both questions were never, and 0 Yes (coded 1) – if the response was 1 or more on both questions.

2.6. Individual-level factors

Emotional wellbeing during Menstruation were measured using six items from the Questionnaire Assessing Girls’ Menstrual Hygiene Practices in East Africa (Irise International, 2013). Participants were asked to answer (yes/no) if they experienced similar feelings during their period. The sample questions included: “During my menstruation period I feel that I am a failure.”; “During my menstruation period I feel I am no good” with higher score indicating higher levels of poor emotional wellbeing experienced during menstruation (α = 0.6978).

General Physical Health.

Participants were asked to evaluate their physical health on a scale: Very Poor-1, poor, fair, good, excellent-5) by answering the question, “What would you say about your physical health?”. The dummy variable of poor physical health created by using no (0) if they indicated that their physical health was good or excellent and yes (1) – if they indicated otherwise.

Dysmenorrhea

(pain with menstruation), participants were asked to report on a scale of 0 (pain-free) to 10 (worst pain I have ever experienced), the average level of pain they experienced during their menstruation period.

Personal financial constrains –

the adolescent girls were asked if they agree on the statement “I do not have enough money to buy disposable sanitary pads from a shop”: Yes was coded as “1” and No as “0”.

Menstruation Knowledge.

The level of knowledge, readiness and myths about menstruation among girls were assessed by three items: 1) “Menstruation is a disease,” 2) “Pain during the menstruation period means that someone is sick,” and 3) “Menstrual blood contains dangerous substances.” All respondents were asked to indicate whether specific statements about menstruation were true (coded as 1) or false (coded as 0). (α = 0.6341).

2.7. Family level factors

Family Cohesion was measured using seven items from the Family Environment Scale (FES) (Moos & Moos, 1994). The scale measures the degree of commitment, help, and support that family members provide for one another. Respondents were asked to rate how often each item occurred in their family, on a 5-point Likert scale, with 1 = never and 5 = always, with higher scores indicating higher levels of family cohesion. The theoretical range for this scale is 7–35, with high scores indicating higher levels of family cohesion (α = 0.7291).

Family relationships and perceived support was assessed by using six items from the Friendship Qualities Scale (Bukowski et al., 1994) and 17 items from the Social Support Behaviors (SS-B) Scale (Vaux et al., 1987). Respondents’ responses were rated on a 5- point Likert scale, with 1 = never and 5 = always. Family support was measured using 6 items related to things that parents/caregivers sometimes do with their children. Respondents were asked to rate how often each item occurred in their family, on a 5-point Likert scale, with 1 = never and 5 = always. The theoretical range for the merged scale is 23–115, with high scores indicating higher levels of family relationships and support (α = 0.7832).

Family Communication was assessed using the Family Communication scale (Olson et al., 2004). Respondents were also asked to rate how comfortable they felt talking to their caregivers about specific topics. Responses were rated on a 4-point scale, with 1 = very uncomfortable and 4 = very comfortable (α = 0.8476).

Family Sufficiency/Poverty Indicators.

Family sufficiency was measured using four items related to the adolescents’ experience in the previous three months. The questions were: “Over the past 3 months, how often have you gone have you gone without enough food to eat?; “Over the past 3 months, how often have you gone without enough clean water?”; “Over the past 3 months, how often have you gone without medicine?”; “Over the past 3 months, how often have you gone without school expenses for fees?”. Respondents were asked to rate how often each item occurred in their family, on a 5-point Likert scale, with 1 = never and 5 = always. Higher scores indicate a high level of poverty (α = 0.6894).

2.8. School-level factors

Teachers Support and Classmates support were measured quantitatively using corresponding subscales from the Friendship Qualities Scale (Bukowski et al., 1994) and additional items adapted from the Social Support Behaviors (SS-B) Scale (Vaux et al., 1987). The scale assesses the impressions of the quality of girls’ relationships with their teachers and classmates. Respondents’ responses were rated on a 5-point Likert scale, with 1 = never and 5 = always (α = 0.6987).

Gender norms about schooling/education. Gender stereotypes and education were measured by four statements from the Attitudes Towards Women Scale for Adolescents (Galambos et al., 1985). Statements include “On average, girls are as smart as boys.”, “More encouragement in a family should be given to sons than daughters to go to college.”, “It is more important for boys than girls to do well in school.”, and “Boys are better in school than girls.” The girls were asked to indicate whether they agreed with each statement related to how men and women act. Items had a “Yes” coded as “1” or “No” coded as “0” (α = 0.3396).

Adequacy of school resources (WASH Facilities) was assessed by two items from the Questionnaire Assessing Girls’ Menstrual Hygiene Practices in East Africa (Irise International, 2013), related to reasons for missing school: 1) “I miss school during my menstruation period because there is not anywhere for girls to wash and change at school” and 2) “I miss school during my menstruation period because there is nowhere to dispose of sanitary products.” These items were self-reported by the participants, thus capturing their personal experiences and perceptions of the WASH facilities available at their school.

2.9. Community-level factors

Menstruation Taboos were assessed through direct self-report by participants using two items from the Questionnaire Assessing Girls’ Menstrual Hygiene Practices in East Africa (Irise International, 2013). Participants were asked to indicate whether they believed certain statements related to how society views menstruation: 1) “During menarche, girls should not leave home” and 2) “I miss school during menstruation period because of cultural or religious reasons.” These items reflect commonly held cultural beliefs that contribute to school absenteeism by limiting girls’ mobility and participation during menstruation (α = 0.6341).

Community Resources and Poverty.

Distance to schools, distance to medical/health centers, distance to clean water sources and existence of bank in the community were used as proxy measures for community poverty.

2.10. Data analysis

Participants were clustered in schools. First to account for the multilevel structure of the data, we ran a likelihood ratio test that compared the model fit of an unconditional multilevel logistic regression model to that of a single-level unconditional logistic regression model (Snijders & Bosker, 2011). We used the intraclass correlation (ICC), the proportion of variance in the outcome variable explained by the hierarchical model’s grouping structure to assess the data’s hierarchical character (Snijders & Bosker, 2011). The low ICC suggested that school absenteeism differed far more between individual adolescent girls than across schools. Therefore, in the following steps, a nested logistic regression analysis was used to test the significance of blocks of predictors.

Using the baseline model comprising only of individual factors (model 1), the regression model was built adding one block (family, school, and then community level factors) at a time to the baseline model to form the full model. We estimated the odds that an event (absence from school because of menstruation) would occur while adjusting for clustering of observations at the school level (using marginal models) to obtain robust standard errors. Model fit was evaluated using Bayesian Information Criterion (BIC) (Stone, 1977) and Akaike’s Information Criterion (AIC) (Akaike, 1974). The model with smaller AIC and BIC indicates a better fit. The residual analysis of the final model and Hosmer-Lemeshow (HL) tests were also used to assess the model fit (Long & Freese, 2006). Statistical significance was determined at the 5 % level. Data cleaning, management and all analysis were conducted in STATA 15.

3. Results

A total of 1,123 girls out of 1260 girls who indicated that they had started menarche were included in the final analysis. 424 (37.76 %) of the adolescent girls reported school absenteeism during their menstruation.

The descriptive analysis (Table 1) showed that in regards of individual risk factors, about 63 % of participants reported feeling pain associated with period on the scale from 0 to 10 (median = 3, IQR = 6). Absolute majority of the girls (97 %) also reported some concerns about their emotional wellbeing during menstruation (median = 3, IQR = 2, Range = 0–6) and 62.87 % reported financial constrains to buy disposable pads. In terms of family-level factors, participants reported high levels of family cohesion (median = 35, IQR = 11), family support (median = 24, IQR = 6) and family communication (median = 26, IQR = 8). More than 80 % of participants reported issues with poverty (median = 6, IQR = 5).

Table 1.

Description of the Sample (n = 1,123).

Characteristic	Number/ Median	Percent (%)/IQR
Missing school because of menstruation (DV)	424	37.76
Individual Factors
Emotional well-being during menstruation (Range = 0–6)	3	2
Poor/fair Physical health	200	17.81
Pain during menstruation (Range = 0–10)	3	6
Period Poverty	706	62.87
Family Factors
Family Cohesion (Range 9–45)	35	11
Family Support (Range = 9–30)	24	6
Family Communication (Range = 11–44)	26	8
Family Poverty (Range = 4–20)	6	5
Main Caregiver: Female (mother, grandmother, aunt)	458	40.78
School Factors
Poor School infrastructure	712	63.40
Supportive Teachers (Range = 10–30)	23	6
Supportive Classmates (Range = 7–25)	18	5
Gender (bad) norms related to schooling/and education	960	85.49
Community Factors
No menstruation Taboos	617	54.94
Distance to clean water source: near	941	83.79
Distance to school: near	636	56.63
Bank in community	401	35.71
Distance to medical/health center: near	909	80.94
Demographics of the sample	mean (SD)
Age (Range: 14–17)	15.37 (0.87)
Orphanhood status
Double orphan	24	1.90
Single orphan	191	15.16
Non-orphan	1045	82.94
Primary caregiver
Biological parent	965	76.59
Grandparent	140	11.11
Other parent	155	12.30
Number of people in household size (Range = 2–31)	7.00 (2.71)
Number of Children in a household size (Range 0–13)	3.50 (2.10)
Asset ownership (Range: 0–20)	11.46 (3.26)
Missing school during menstruation due to cultural or religious reasons
Yes	201	17.90
No	922	82.10
Menstrual knowledge (Range: 0–9)	3.52 (1.93)
Self-esteem (Range: 16–40)	34.01 (4.57)

In terms of school-level factors, a 63.40 % of participants reported inadequate infrastructure (no water and no place to dispose the pads) at their schools.

Finally, regarding community-level factors, 45 % of the participants supported the notion that girls should not leave their houses and/or reported doing so because of their religious traditions. The analysis revealed that majority of the girls were not only poor, but they also lived in poor communities as well as only 56.63 % reported having school near (within 2 kms) (Table 1).

Following the descriptive analysis, we tested the multilevel structure of the data as the girls were nested in schools. We examined a likelihood ratio test that compared the model fit of an unconditional multilevel logistic regression model to that of a single level unconditional logistic regression model. This test was not statistically significant (p = 0.518). Also, the intraclass correlation (ICC) of the unconditional logistic model was very low (0.0197), suggesting that school absenteeism differed far more between individual adolescent girls than across schools.

To test H₁, we tested the null hypothesis that the cumulative effects of individual, family, school or community factors were 0. Based on the likelihood-ratio tests, we rejected the null hypothesis that cumulative effects of different blocks were 0 (Table 2). Specifically, the analysis showed that all blocks of risk factors (Individual, Family, School, and Community) had a significant contribution to adolescent girls’ school absenteeism during menstruation. With the addition of blocks, the model fit improved based on the lower values of the Bayesian Information Criterion (BIC) and Akaike’s Information Criterion (AIC) (Table 2). The residual analysis did not reveal influential observations and Hosmer-Lemeshow (HL) test (chi2(1102) = 1123.79, p = 0.3173) showed a good model fit.

Table 2.

Comparing the models: measures of fit.

Nested Models	LL	LR	df	Pr > LR	AIC	BIC
Model (1) with Individual Factors	−716.4255	55.93	5	0.0000	1444.851	1474.994
Model (2) with Individual and Family Factors	−683.6685	65.51	6	0.0000	1391.337	1451.622
Model (3) with Individual, Family and School Factors	−667.001	33.33	4	0.0000	1366.002	1446.382
Full Model (4) with Individual, Family, School and Community Factors	−658.5038	16.99	5	0.0045	1359.008	1464.507

^*LL – Log pseudolikelihood; LR – Likelihood Ratio; df- degree of freedom, AIC – Akaike’s Information Criterion; BIC- Bayesian Information Criterion

Individual Factors:

In the analysis investigating hypothesis 2, the results showed that in model (1) which comprised only blocks of individual factors, adolescent girls who experienced financial constrains to buy disposable sanitary pads were 34 % more likely to miss school than girls not affected by this financial constrain (OR: 1.34, p = 0.017). However, the statistically significant effect was lost in models 2, 3, 4 with the addition of family, school and community factors.

The same trend was observed for girls experiencing poor emotional wellbeing during menstruation. They had increased odds of school absenteeism because of menstruation (by 10 %). However, this became non– significant, when school and community factors were introduced into the model (Table 3, Model 3 & 4).

Table 3.

Results from Nested Logistic Regression.

Variable	Odds Ratio	95 % CI		Robust Std. Err.	p- value	Log pseudolikelihood (LL)	Wald chi2 (11)	Pseudo R2
		LL	UL
Model (1) with Individual Factors						−716.43	47.00	0.04
Constant	0.27	0.17	0.42	0.06	0.00
Menstruation knowledge	0.98	0.89	1.07	0.05	0.60
Poor emotional wellbeing associated with menstruation	1.11	1.03	1.19	0.04	0.01
Poor general physical health	1.02	0.73	1.44	0.18	0.90
Dysmenorrhea (Pain associated with menstruation)	1.10	1.06	1.15	0.03	0.00
Personal Financial Constrains	1.34	1.05	1.71	0.16	0.02
Model (2) with Individual and Family Factors						−683.67	83.12	0.08
Constant	4.64	1.75	12.30	2.31	0.00
Menstruation knowledge	1.01	0.92	1.11	0.05	0.88
Poor emotional wellbeing associated with menstruation	1.10	1.02	1.18	0.04	0.01
Poor general physical health	0.89	0.64	1.25	0.15	0.51
Dysmenorrhea (Pain associated with menstruation)	1.09	1.04	1.14	0.02	0.00
Personal Financial Constrains	1.16	0.89	1.50	0.15	0.28
Family cohesion	1.00	0.98	1.02	0.01	0.86
Family support	0.93	0.90	0.97	0.02	0.00
Family communication	0.97	0.95	0.99	0.01	0.00
Family sufficiency	0.82	0.74	0.89	0.04	0.00
Female caregiver	1.16	0.94	1.43	0.12	0.16
Model (3) with Individual, Family, and School Factors						−667.00	167.16	0.10
Constant	10.14	3.51	29.25	5.48	0.00
Menstruation knowledge	1.00	0.90	1.10	0.05	0.94
Poor emotional wellbeing associated with menstruation	1.05	0.97	1.13	0.04	0.21
Poor general physical health	0.91	0.65	1.27	0.16	0.58
Dysmenorrhea (Pain associated with menstruation)	1.08	1.03	1.13	0.03	0.00
Personal Financial Constrains	1.09	0.83	1.43	0.15	0.52
Family cohesion	1.00	0.98	1.03	0.01	0.64
Family support	0.95	0.91	0.99	0.02	0.01
Family communication	0.98	0.96	1.00	0.01	0.02
Family sufficiency	0.85	0.77	0.94	0.04	0.00
Female caregiver	1.19	0.96	1.48	0.13	0.12
School WASH infrastructure	0.67	0.56	0.80	0.06	0.00
Supportive teacher	0.95	0.92	0.99	0.02	0.01
Supportive classmates	1.00	0.97	1.04	0.02	0.79
Gender norms related to school/education	1.16	1.03	1.30	0.07	0.02
Full Model (4) with Individual, Family, School and Community Factors						−658.50	167.16	0.10
Constant	10.66	3.26		6.44	0.00
Menstruation knowledge	1.01	0.91		0.05	0.92
Poor emotional wellbeing associated with menstruation	1.03	0.95		0.04	0.47
Poor general physical health	0.91	0.64		0.17	0.62
Dysmenorrhea (Pain associated with menstruation)	1.08	1.03		0.03	0.00
Personal Financial Constrains	1.07	0.81		0.15	0.64
Family cohesion	1.00	0.98		0.01	0.73
Family support	0.95	0.92		0.02	0.01
Family communication	0.98	0.96		0.01	0.02
Family sufficiency	0.86	0.78		0.04	0.00
Female caregiver	1.20	0.97		0.13	0.10
School WASH infrastructure	0.71	0.60		0.06	0.00
Supportive teacher	0.95	0.92		0.02	0.02
Supportive classmates	1.00	0.97		0.02	0.88
Gender norms related to school/education	1.13	1.00		0.07	0.05
Bank in a community	0.72	0.55		0.10	0.02
Distance to medical /health facility: near	0.81	0.62		0.11	0.12
Distance to clean water: near	1.10	0.78		0.19	0.59
Cultural taboos related to menstruation	0.85	0.55		0.19	0.47

Note. CI – confidence Interval; LL- Lower limit; UL = Upper limit.

^*bold font represents statistically significant findings (p < 0.05).

On the contrary, dysmenorrhea (painful period) remained significant in all 4 models (Table 3). The analysis showed that girls who experienced painful period were 8 % more likely to miss school while adjusting for all other factors in the model. The analysis also showed that increasing pain by 1 SD (3.6 point on the scale) was associated with 31.9 % increased odds of school absenteeism (Table 3).

Family Factors:

In terms of family factors, family support and family communication were both associated with school absenteeism. Specifically, a 1 unit increase in family support was associated with reduced odds of school absenteeism by 4.9 % (p = 0.01). Similarly, a 1 unit increase in family communication was associated with a 2.2 % (p = 0.024) reduced odds of school absenteeism while adjusting for all other variables. The analysis also showed that with each unit increase in family sufficiency scores (which equates to lower levels of poverty), the girls’ odds of school absenteeism during menstruation decreased by 13.8 %.

School Factors:

The exploration of school-level factors showed both adequate infrastructure and supportive teachers as potentially having a positive impact on adolescent girls’ school attendance. Specifically, adequate school WASH infrastructure was associated with a statistically significant reduction in girls school absenteeism by 28.6 % (p = 0.000), accounting for all other factors. Similarly, with each unit increase in support from teacher, the odds of absenteeism decreased by almost 5 % (p = 0.018) when holding all other variables at constant. The results also indicate that school proximity to the girls’ homes helped to reduce school absenteeism by 28.1 % (p = 0.016) in the fully adjusted model.

Community Factors:

Among community factors, cultural taboos related to menstruation showed the strongest impact on girls’ school absenteeism. With each unit increase in cultural taboos, there was 34.3 % increased odds of school absenteeism during menstruation, in the fully adjusted model (Table 3, Model 4).

The residual analysis of the final model and the HL tests showed a good fit of the final model. Model fit was also tested according to the AIC and BIC criteria and we compared models while adding a block of covariates. Model 4 showed a better fit also based on AIC and BIC criteria (Log likelihood = −658.5038, df = 5, AIC = 1359.008, BIC = 1464.507).

4. Discussion and conclusion

This study aimed to investigate the blocks of individual, family, school, and community level factors associated with school absenteeism of adolescent girls during their menstruation. Our analysis showed that blocks of individual, family, school, and community-level factors all significantly contributed to adolescent girls’ school absenteeism. With these blocks, analysis of discrete factors showed that some factors appeared more important than others.

Within individual-level factors, adolescent girls who reported pain associated with menstruation (dysmenorrhea) were more likely to report missed school days due to menstruation. This finding is consistent with another study conducted in Entebbe, Uganda that also found that menstrual pain to be a key factor in adolescent girls’ school attendance (Miiro et al., 2018). It is worth to mention that menstrual pain remained the only statistically significant predictor of school absenteeism among individual level factors, pointing to the direct relationship between physical health and school absenteeism (Hennegan et al., 2019).

Within family-level factors, adolescent girls with higher perceived family support, higher levels of family economic sufficiency, and who were more comfortable talking to their caregivers, were less likely to miss school due to menstruation. The result related to family sufficiency points to the importance of economic empowerment at the family level. Specifically, the lack of menstrual hygiene products is one of the major contributing factor to school absenteeism during menstruation (Dolan et al., 2014; Greed, 2014). Even though in our analysis financial constrains as an individual factor was not statistically significant, it is possible that families with higher economic sufficiency are more supportive and girls tend to miss less school days with adequate MHM (Tegegne & Sisay, 2014).

Adolescent girls who felt more comfortable talking to their caregivers about sensitive topics, such as puberty were less likely to miss school. It could be that they were more likely to discuss and receive information about the aspects of menstruation and MHM. This may enhance adolescent girls’ confidence and ability to practice effective and safe MHM both at home and at school (Blake et al., 2018). Indeed, Miiro and colleagues (Miiro et al., 2018) documented that lack of knowledge of menstruation in adolescent girls is largely due to caregivers’ disengagement regarding sexual and reproductive health topics. Similarly, Tegegne and Sisay (Tegegne & Sisay, 2014) in their Ethiopia study, attributed the knowledge gaps related to menstruation and its management among adolescent girls to low parent–child communication about MHM. As such, caregivers who are perceived to be more supportive and whom adolescents feel comfortable talking to, may hold positive norms or be better able to negate stigma and taboos related to menstruation (Day, 2018; Kumar & Srivastava, 2011).

Based on the integrated model of menstrual experience, several pathways can potentially explain the relationship between social support – both within the family and school, and school absenteeism in the context of MHM (Hennegan et al., 2019). Social support can help to mediate shame and distress associated with menstruation; enhance confidence to manage menstruation effectively; shape positive perceptions as well as good practices of MHM (Hennegan et al., 2019), all of which would positively impact school attendance among adolescent girls.

Our study also showed that having adequate WASH infrastructure at school was significantly associated with lower odds of school absenteeism. The negative impact of inadequate WASH infrastructure on school attendance, particularly during menstruation, has been a common theme in previous studies across SSA (Boosey et al., 2014; Crofts & Fisher, 2012; Dolan et al., 2014; Greed, 2014; Sommer et al., 2015; Sumpter & Torondel, 2013; Sumpter & Torondel, 2013). In Uganda, a study among 13–16-year-old school-going girls in Rukungiri district found that lack of a place for girls to wash and change at school was the commonly cited reason for school absenteeism (Boosey et al., 2014). Similarly, in Southern Uganda, difficulties in MHM were attributed to inadequate water supply to bath or wash used menstrual hygiene products; lack of private and hygienic facilities to change; and non-existent disposal containers (Crofts & Fisher, 2012). This finding point to the urgent need to improve sanitation facilities within schools (Crofts & Fisher, 2012; Kuhlmann et al., 2017; Lahme et al., 2018; Sommer, 2010). Specifically, adolescent girls (in previous studies) have recommended adequate sanitary facilities should include functioning cubicle doors and roofs in wash places for privacy; water and soap basins in the same place, better lighting and disposable bins closer to the sanitary facilities (Boosey et al., 2014; Miiro et al., 2018).

At the community level, distance to school and cultural taboos related to menstruation remained significant predictors of school absenteeism within the block of community factors. Adolescent girls who lived closer to their school were less likely to report school absenteeism during menstruation, compared to those who traveled longer distances. In the context of MHM, shorter distance to the school could possibly afford girls the option to return home during school breaks, given that home is considered a solace that may offer better privacy or hygiene to manage menstruation (UNICEF, 2015). Additionally, in cases where transportation is either too costly or unavailable, shorter walking distance may be less burdensome on days where girls experience heavy menstrual flows (Lahme et al., 2018). They also pointed out that walking long distances with cloths or rags as a pad can cause friction and to avoid this discomfort, girls would likely be less motivated to attend school (Lahme et al., 2018).

In addition, cultural taboos related to menstruation were found to be significantly associated with increased odds of school absenteeism among adolescent girls. Existing literature underscores the pivotal role of cultural context and upbringing in shaping menstrual knowledge, perceptions, and attitudes (Marván et al., 2017; McMahon et al., 2011; Öztürk & Güneri, 2021; Wong, 2011). Although this study did not directly assess broader community attitudes toward menstruation, participants’ responses suggested internalization of cultural and religious norms, which contributed to self-imposed restrictions and school absenteeism during menstruation. Prior research has documented a range of cultural norms, including the association of menstruation with sexuality (Dolan et al., 2014; Sumpter & Torondel, 2013) and stigma, shame, and social isolation (Maulingin-Gumbaketi et al., 2022; Sommer, 2009). In certain contexts, beliefs surrounding menstrual impurity and taboos against the visibility of menstrual blood impose additional social restrictions, further contributing to school absenteeism (Hennegan et al., 2019; Sumpter & Torondel, 2013). A lack of accurate menstrual hygiene management (MHM) information further diminishes girls’ confidence and capacity to manage menstruation in public environments, including schools (Kuhlmann et al., 2017; Lahme et al., 2018). To effectively address these barriers, integrated interventions, comprising menstrual health education, community sensitization, and improved access to menstrual products, have been recommended (Hennegan et al., 2016, 2022; Hennegan & Sol, 2019).

4.1. Limitations and strengths

Study findings should be interpreted with a few limitations in mind. This was a cross-sectional study; thus, it is difficult to establish a causal relationship between the multi-level predictors and school absenteeism during menstruation. Even though focusing on adolescent girls living within a family was relevant to the study objectives that sought to examine multi-level determinants, including those specific to the family, we cannot generalize our findings to adolescent girls who are not living with families. All data were based on self-report, which is subject to social desirability and recall bias. Additionally, the data were collected eight years ago. Although many challenges related to menstruation and education likely remain relevant, changes in policy, programming, or social norms over time may influence the current applicability of the results.

Even with these limitations, this study has several strengths. To our knowledge this is the first study in SSA that has used the ecological approach which allowed an exploration of multilevel drivers of school absenteeism among adolescent girls during menstruation. This provides an opportunity to identify intervention points, if targeted simultaneously, could potentially reduce school absenteeism among adolescent girls. Utilizing theoretical frameworks that are specific to menstrual experiences ensured rigor in the selection of variables that are relevant to the study objectives and guided the research analysis. One of the questions used for the outcome variable assessed frequency of school absenteeism due to menstruation which enhances the ability to capture patterns of absenteeism related to menstruation. This study had a relatively large sample size (n = 1,223) which enhances the precision of the estimates of the association between the predictor variables and the outcome (Phillips-Howard et al., 2016). Importantly, the study’s finding reinforces recommendations from other SSA countries aimed at improving knowledge, awareness and access to pain relief methods to adolescent girls to help increase school attendance and graduation rates (Crofts & Fisher, 2012; Miiro et al., 2018; Sommer, 2010).

4.2. Practice implications

The findings of this study contribute to the growing evidence for the need for multi-faced and context-specific interventions and provide clear guidance for strengthening school attendance during menstruation among adolescent girls in LCIMs (Hennegan et al., 2022; Hennegan & Sol, 2019; Ssesanga et al., 2024; Stoilova et al., 2022; Swe et al., 2022). Menstrual pain remained statistically significant across all models, indicating a critical need for interventions that address pain management among adolescent girls. Schools and health partners should integrate both pharmacological and non-pharmacological pain relief strategies into menstrual health education and make these options more accessible to students. Programs should consider incorporating caregiver engagement components that encourage open, supportive dialogue about menstruation in the home, alongside efforts to improve girls’ menstrual health knowledge. Cultural taboos showed the strongest association with increased absenteeism. These results support the implementation of culturally informed community sensitization programs that challenge restrictive norms and provide accurate information about menstruation. Engaging parents, religious leaders, and community influencers is essential to reducing stigma and promoting menstruation as a healthy and manageable part of adolescent development.

4.3. Further research

This study identified several factors that merit deeper exploration in future work. Menstrual pain was a consistently significant predictor of absenteeism, yet few studies have evaluated scalable, school-based pain management interventions. Research should examine the feasibility and outcomes of integrating pain relief strategies into school health services, especially in resource-constrained settings. Family dynamics also deserve further investigation. While quantitative findings show that family support and communication reduce absenteeism, more qualitative research is needed to understand how these dynamics function and how communication about menstruation is shaped by household norms, parental attitudes, and gender roles.

5. Conclusion

The present study found that the comorbidity of individual, family, school, and community-level factors was associated with school absenteeism during menstruation among adolescent girls in Southern Uganda. Based on these findings, multifaceted interventions are needed to address not only physical symptoms associated with menstruation but also to enhance support and improve communication related to MHM within families and communities, while addressing cultural taboos and stigma around MHM to prevent school absenteeism among adolescent girls during menstruation. The preliminary results of the pilot MENISCUS (Menstrual Health Interventions, Schools and Communities) initiative, which provides a multi-component, school-based program combining puberty education, menstrual product provision, WASH improvements, teacher training, pain management, and menstrual-health action groups, demonstrated mixed findings on mental health problems, educational performance, and menstrual health (Kansiime et al., 2020; Nelson et al., 2025). Further research is needed to assess the comprehensive effects of similar integrated multifaceted interventions to reduce menstrual-related absenteeism and support education equity.

Data availability

Data will be made available on request.