A decade of retail purchasing data revealed a hidden pattern of menstrual pain across England, suggesting that people in lower-income communities may face greater barriers to managing period-related pain.
Study: What can shopping transactional data reveal about relative prevalence of menstrual pain and period poverty in England? Image credit: Halfpoint/Shutterstock.com
Period pain affects about 91% of people who menstruate, but little is known about its extent and determinants. A recent study in the journal PLOS Digital Health tests the use of a novel proxy tool based on shopping data in examining the prevalence of menstrual pain and exploring whether purchasing patterns might reflect inequalities linked to period poverty in England.
The unrecognized problem of menstrual pain
Menstrual pain or dysmenorrhea is both common and can severely impact quality of life, limit ordinary activity, and promote absenteeism. However, research on this topic is limited, especially at the population level. Social and medical attitudes have often framed menstrual pain as something to be endured as part of being female, rather than treated as an illness, albeit recurrent and temporary.
Medical help is not always sought, either due to such a mindset, questions about treatment, or barriers to accessing care. The stigma attached to menstruation often accompanies the pain and disability linked to menstrual pain as well, discouraging them from seeking help, increasing their emotional burden, and possibly leading to ill-health.
Tracking menstrual health through shopping behavior
Public health policy has been slow to recognize and act on these observations due to small study sizes in prior research. In this study, the authors highlight the value of large-scale datasets in filling this gap and hope to provide a new tool for population-level menstrual health research.
The study used mass transactional data from a major health and beauty retailer’s shopping logs to explore behavioral patterns potentially linked to menstrual pain and possible period poverty across time and different regions of England.
The shopping data was retrieved from a leading health and beauty retail chain, with deprivation based on publicly available Indices of Multiple Deprivation (IMD) data compiled by the UK Office for National Statistics (ONS), which tracks neighborhoods across the UK.
Menstrual and pain relief purchases
Using more than 211 million unique loyalty-card transactions collected over ten years, between April 2006 and April 2015, the researchers developed a behavioral proxy for menstrual pain. They looked only at purchases involving tampons or sanitary pads and oral or topical pain relief products.
They defined a transaction involving menstrual pain relief as a shopping basket that contained both menstrual products and pain-relief medications. They acknowledge the indirect nature of this data, but argue that the strong purchasing patterns strengthen confidence in the proxy measure.
Among the 18% of loyalty-card customers who bought menstrual products, there were an average of 126 menstrual transactions per customer over the study period. This indicates a mean spend of £1453. Among the 20 most frequently purchased menstrual products, pads accounted for 60% and tampons for 40%.
Pain purchases were made by 31% of this customer base. On average, these customers made 354 pain-related transactions per customer at a mean total spend of £1279 over the study period. However, the data suggest customers may be more price-sensitive when purchasing pain products than menstrual products, which may influence decisions about pain relief.
Co-purchasing patterns suggest widespread pain management
Overall, 26.7% of customers who bought menstrual products also purchased pain medication in the same basket, suggesting that menstrual pain-related purchases are common.
The authors note that this likely underestimates the true burden because many people may manage pain differently or may not be able to afford medication. The researchers describe their measure as a likely lower-bound estimate of menstrual pain-related purchasing behavior because it would not capture people who use prescription medications, hormonal contraceptives, or heat-based therapies, or those who purchase menstrual and pain-relief products separately. Notably, those who bought both product types had more transactions and higher total spend than those who bought either menstrual or pain-relief products alone.
Among those who have ever bought menstrual and pain products at the same time, half the menstrual purchases included a pain product. This is a fourfold increase over the chance they would buy a pain product on its own.
Shopping data mirrors the 28-day cycle
To validate their approach, the researchers looked for regular patterns in menstrual product purchasing behavior. The most common interval between purchases was 28 days, matching both the mean and median intervals observed within a typical menstrual cycle. This closely matches the average menstrual cycle.
The researchers thus demonstrated that large-scale transactional data can reflect a biologically relevant 28-day menstrual cycle signal. The authors argue that this periodic pattern supports using shopping data as a behavioral proxy for menstrual health.
Income strongly predicts menstrual pain purchases
Further analyses showed socioeconomic disparities in menstrual-related purchases. Income emerged as the strongest predictor of menstrual pain purchasing patterns in their machine learning (ML) models.
Regional income was strongly correlated with the number of transactions in all three areas: menstrual, pain, and menstrual pain. Thus, areas with higher deprivation scores generally had higher overall transaction rates but bought items in these three categories at lower rates.
Notably, customers in regions with the lowest incomes relative to the national average were 32% less likely to purchase menstrual pain products than those in regions with the highest mean income. The researchers suggest that this may reflect differences in affordability, access, or purchasing behavior rather than differences in menstrual pain itself, and could indicate socioeconomic inequalities in menstrual health management.
Single-retailer dataset limits broader conclusions
The study acknowledges important limitations. The data came from a single retailer and included only loyalty-card users, which could limit its generalizability to the broader population. Also, the study findings are based on probabilistic estimates and not direct measurements.
However, pain products bought with menstrual products need not be for menstrual pain; they could be for other family members or part of a stock-up. Women might also be buying pain items elsewhere or at non-period times.
Potential menstrual health inequalities
Overall, the study concludes that shopping transaction data may offer a novel method for monitoring menstrual pain and investigating socioeconomic inequalities in menstrual health management at a national scale. The findings suggest that socioeconomic deprivation could affect the experience and management of menstrual pain.
This result should motivate further study to establish whether this variation in menstrual pain transactions is due to the inability of those in more deprived areas to afford period pain medication, serving as empirical evidence health inequalities in England.
The authors suggest that the findings could help inform public health programs and workplace and educational policies that recognize and accommodate menstrual pain-related challenges, particularly among economically disadvantaged groups.
Download your PDF copy now!
Breast reduction surgery is linked to lower diabetes and heart risk