MPH · MSc · Public health dissertations
50 research questions matched with a study design, a statistical method, a free dataset and a related university thesis. Move from a topic idea to a workable research plan.
50 Research questions
10 Sub-fields
10 Dataset sources
3 Universities
2022–26 Thesis years
By Siddharth Gupta · Statssy
Most topic lists give you a title and then leave you alone. That is where students get stuck.
In my 12 years of helping MPH and MSc students, the topic was rarely the real problem. The problem came two months later. They had a topic but no data, and no idea which test to run.
So this list works differently. Every question below comes with three things you need on day one: a study design that fits, a statistical test appropriate for the objective, and a free dataset to investigate. Each also links to a real, related thesis.
How to use this list Pick two or three questions from the sub-field you care about. Open the dataset link and check its codebook. Confirm your outcome and main variables exist. Read the related thesis, paying particular attention to its methods chapter. Narrow your question to one population, place and period. Take the shortlist to your supervisor or discuss it with me . Jump to a sub-field 01. Infectious disease and outbreaks 02. Non-communicable diseases 03. Maternal and child health 04. Mental health 05. Nutrition and obesity 06. Health policy and health systems 07. Environmental health 08. Health inequalities and social determinants 09. Vaccination and immunisation 10. Digital health and data science Research area 01 / 10
Infectious disease and outbreaks Outbreak topics suit count models and time series. Country-level data is free and updated often.
TOPIC 01 Do warmer and wetter years bring more dengue cases across tropical countries? Study design Ecological panel Statistical test Negative binomial panel regression Free dataset GBD Results + World Bank WDI TOPIC 02 Does forest loss explain rising vector-borne disease burden across countries? Study design Ecological panel Statistical test Fixed effects panel regression Free dataset GBD Results + World Bank WDI TOPIC 03 Has the diarrhoeal disease burden in older adults fallen as fast as in children? Study design Trend analysis Statistical test Joinpoint regression Free dataset GBD Results TOPIC 04 Does armed conflict raise cholera cases once safe water access is accounted for? Study design Ecological panel Statistical test Fixed effects regression with an interaction term Free dataset WHO GHO + World Bank WDI TOPIC 05 Can COVID-19 case counts warn of hospital admissions one to two weeks ahead? Study design Time series Statistical test Distributed lag model Free dataset OWID COVID-19 Data Research area 02 / 10
Non-communicable diseases NCD topics are the easiest place to start. Large national surveys give you blood pressure, glucose and lifestyle data for free.
TOPIC 06 Do hypertension screening rates differ by insurance status and race in US adults? Study design Cross-sectional Statistical test Survey-weighted logistic regression Free dataset BRFSS TOPIC 07 Does emotional support help adults stay on their blood pressure medicine? Study design Cross-sectional Statistical test Logistic regression Free dataset BRFSS TOPIC 08 Which Indian adults have high blood pressure but have never been told so? Study design Cross-sectional Statistical test Multilevel logistic regression Free dataset NFHS-5 India TOPIC 09 Who is most likely to have diabetes that has never been diagnosed? Study design Cross-sectional Statistical test Survey-weighted logistic regression Free dataset NHANES TOPIC 10 Do adults eat and move differently after a type 2 diabetes diagnosis? Study design Cross-sectional comparison Statistical test Propensity score matching with linear regression Free dataset NHANES Research area 03 / 10
Maternal and child health This is the richest area for free data. DHS, NFHS and MICS were built for these questions.
TOPIC 11 Do women with chronic conditions have higher odds of a low birth weight baby? Study design Cross-sectional Statistical test Logistic regression Free dataset NFHS-5 India TOPIC 12 What decides whether a mother seeks care when her child has diarrhoea? Study design Cross-sectional Statistical test Multilevel logistic regression Free dataset DHS Program TOPIC 13 Do maternal health conditions shorten exclusive breastfeeding? Study design Cross-sectional Statistical test Cox regression on breastfeeding duration Free dataset NFHS-5 India TOPIC 14 Which causes drive maternal deaths, and how have they shifted since 2000? Study design Trend analysis Statistical test Joinpoint regression Free dataset GBD Results TOPIC 15 Do more antenatal visits lower the odds of low birth weight in poorer households? Study design Cross-sectional Statistical test Logistic regression with an interaction term Free dataset UNICEF MICS Research area 04 / 10
Mental health Mental health topics work well with survey data. Most use ordinal or logistic models and often test moderation.
TOPIC 16 Is there a dose-response link between childhood adversity and adult anxiety? Study design Cross-sectional Statistical test Logistic regression with a dose-response trend Free dataset BRFSS TOPIC 17 Which children with functional difficulties miss out on early learning support? Study design Cross-sectional Statistical test Logistic regression Free dataset UNICEF MICS TOPIC 18 Does emotional support protect sexual minority adults from depression? Study design Cross-sectional Statistical test Logistic regression with moderation Free dataset BRFSS TOPIC 19 Is the burden of depression higher in countries facing conflict and displacement? Study design Ecological panel Statistical test Fixed effects panel regression Free dataset GBD Results + World Bank WDI TOPIC 20 How much of the variation in poor mental health days sits at state level, not individual level? Study design Cross-sectional Statistical test Multilevel negative binomial regression Free dataset BRFSS Research area 05 / 10
Nutrition and obesity Nutrition topics often need mediation or count models. India-focused students have excellent district-level data in NFHS-5.
TOPIC 21 Does physical activity explain the link between household income and obesity? Study design Cross-sectional Statistical test Mediation analysis Free dataset NHANES TOPIC 22 Do malaria and child undernutrition cluster in the same Indian districts? Study design Ecological Statistical test Spatial analysis with negative binomial regression Free dataset NFHS-5 India TOPIC 23 Do children who eat school meals have better diet quality? Study design Cross-sectional Statistical test Propensity score matching Free dataset NHANES TOPIC 24 Which households eat the widest variety of fruits and vegetables? Study design Cross-sectional Statistical test Poisson regression Free dataset NHANES TOPIC 25 Which pregnant women take iron and folic acid for 100 days or more? Study design Cross-sectional Statistical test Multilevel logistic regression Free dataset NFHS-5 India See the full 50-topic PDF download section
Research area 06 / 10
Health policy and health systems Policy topics are where good methods stand out. Difference-in-differences and panel models impress examiners.
TOPIC 26 Did Medicaid expansion raise HPV vaccination and cervical screening? Study design Quasi-experimental Statistical test Difference-in-differences Free dataset BRFSS TOPIC 27 Does higher government health spending improve health outcomes in low and middle income countries? Study design Ecological panel Statistical test Fixed effects panel regression Free dataset World Bank WDI TOPIC 28 Did opioid prescribing limit laws change opioid deaths across US states? Study design Quasi-experimental Statistical test Difference-in-differences with negative binomial regression Free dataset GBD Results TOPIC 29 Do countries with more community health workers have lower child mortality? Study design Ecological panel Statistical test Fixed effects panel regression Free dataset WHO GHO + World Bank WDI TOPIC 30 Does high out-of-pocket spending push women away from facility births? Study design Ecological panel Statistical test Fixed effects panel regression Free dataset World Bank WDI Research area 07 / 10
Environmental health Environmental topics pair exposure data with health outcomes. GBD gives you deaths linked to pollution for every country.
TOPIC 31 How many deaths are linked to air pollution, and is the burden rising faster in some regions? Study design Ecological panel Statistical test Panel regression Free dataset GBD Results TOPIC 32 Do babies in homes that cook with solid fuel have lower birth weight? Study design Cross-sectional Statistical test Logistic regression Free dataset NFHS-5 India TOPIC 33 Do countries hit by more floods and droughts see higher child mortality? Study design Ecological panel Statistical test Fixed effects panel regression Free dataset World Bank WDI TOPIC 34 Does safe water and sanitation access reduce diarrhoea in children under five? Study design Cross-sectional Statistical test Multilevel logistic regression Free dataset UNICEF MICS TOPIC 35 Did years with higher PM2.5 exposure bring a larger preterm birth burden? Study design Ecological panel Statistical test Fixed effects panel regression Free dataset GBD Results Research area 08 / 10
Health inequalities and social determinants Inequality topics test whether a gap exists and whether it changes across groups. Interaction terms and concentration indices do the heavy lifting.
TOPIC 36 Does income predict poor self-rated health, and is the gap different by race? Study design Cross-sectional Statistical test Logistic regression with an interaction term Free dataset BRFSS TOPIC 37 Does education explain gaps in cardiovascular health scores among US adults? Study design Cross-sectional Statistical test Survey-weighted linear regression Free dataset NHANES TOPIC 38 Do rural Indian households face larger gaps in health insurance cover? Study design Cross-sectional Statistical test Multilevel logistic regression Free dataset NFHS-5 India TOPIC 39 Are poorer households more exposed to indoor smoke from cooking fuel? Study design Cross-sectional Statistical test Concentration index with logistic regression Free dataset DHS Program TOPIC 40 Does diabetes risk differ between Asian American adults and other adults at the same BMI? Study design Cross-sectional Statistical test Logistic regression with an interaction term Free dataset NHANES Research area 09 / 10
Vaccination and immunisation Vaccine topics have the best free data in public health. WUENIC alone covers every country since 1980.
TOPIC 41 Did countries that vaccinated older people first see fewer COVID-19 deaths? Study design Ecological panel Statistical test Fixed effects panel regression Free dataset OWID COVID-19 Data TOPIC 42 Did the type of first COVID-19 vaccine a country used shape its booster coverage? Study design Ecological Statistical test Multiple linear regression Free dataset OWID COVID-19 Data TOPIC 43 Did pneumococcal vaccine rollout cut pneumonia deaths in children under five? Study design Quasi-experimental Statistical test Interrupted time series Free dataset UNICEF Immunization Data + GBD Results TOPIC 44 What predicts zero-dose children in India? Study design Cross-sectional Statistical test Multilevel logistic regression Free dataset NFHS-5 India TOPIC 45 Has the drop-off between first and third DTP doses narrowed across countries since 2010? Study design Trend analysis Statistical test Panel regression Free dataset UNICEF Immunization Data Research area 10 / 10
Digital health and data science Digital health topics compare access to technology with health behaviour. Machine learning topics need a simple baseline model to beat.
TOPIC 46 Does owning a mobile phone link to higher HIV testing among women? Study design Cross-sectional Statistical test Logistic regression Free dataset DHS Program TOPIC 47 Does internet use link to better HIV and contraception knowledge among Indian women? Study design Cross-sectional Statistical test Logistic regression Free dataset NFHS-5 India TOPIC 48 Does wider internet access link to higher child vaccination coverage across countries? Study design Ecological panel Statistical test Fixed effects panel regression Free dataset World Bank WDI + UNICEF Immunization Data TOPIC 49 Can machine learning spot undiagnosed diabetes better than logistic regression? Study design Prediction model Statistical test Random forest compared with logistic regression (AUC) Free dataset NHANES TOPIC 50 Do rural adults go longer between routine check-ups than urban adults? Study design Cross-sectional Statistical test Ordinal logistic regression Free dataset BRFSS See the full 50-topic PDF download section
Printable research planning resource
Get all 50 topics in one PDF Download the complete 23-page Statssy guide with all 50 research questions, study designs, statistical methods, dataset links and related thesis references. The PDF includes clickable links and Statssy contact details.
How to turn one of these into your own topic Narrow it. Choose a population, one country or state, and one time period.Check the data first. Review the codebook before committing to the question.Match the test to the outcome. Binary outcomes often call for logistic models, counts for count models, and panel data for longitudinal models.Check your sample size. Use an appropriate power analysis; our sample size calculator can help with preliminary planning.Check the literature. Look for similar studies and identify a distinct research contribution.Need a second pair of eyes?
Turn your shortlist into a research plan. Bring your topic ideas to a free 15-minute meeting. We can review your dataset, proposed analysis and feasibility.
Book a free 15-minute call 