1. Proposed Research Title

“The Syndemic of Modernization: Investigating the Nexus between Education, Technological Adoption, Agricultural Livelihoods, and Non-Communicable Diseases in Somalia.”

2. Study Variables

Dependent Variable (Outcome)

• Presence of Non-Communicable Diseases (NCDs): Derived from qsn3_09 (Chronic illnesses).
  • Note: You will need to create a binary variable (1 = Has at least one chronic illness, 0 = None).

Independent Variables (Predictors)

A. Education Variables: * qsn2_05: Ever attended school (Binary). * qsn2_13: Highest level of education completed (Categorical: Primary, Secondary, Tertiary). * qsn2_03/04: Literacy status (Can read/write).

B. Agriculture & Livelihood Variables: * lab11: Engagement in household farming/livestock (Binary). * _v2631: Ownership of agricultural land. * _v2670-v2672: Livestock ownership (Camels, Cattle, Shoats) — representing traditional rural wealth.

C. Technology Variables: * qsn6_04: Personal mobile phone ownership. * qsn6_08: Internet usage in the last 3 months. * qsn6_11: Access to mobile banking (Financial technology).

D. Control/Covariate Variables: * qsn1_05: Age (Crucial for NCD analysis). * qsn1_03: Sex (Male/Female). * ea_type_n: Residence type (Urban, Rural, Nomadic). * qsn3_22 / qsn3_25: Lifestyle risks (Smoking and Qat chewing). * _v2311: Body Mass Index (BMI).

---

3. Data Preparation (Stata Commands)

You requested code using replace and focusing on the sihbs dataset. Below are the steps to clean and generate your analysis-ready variables.

* 1. Load the dataset
use "SIHBS.dta", clear

* 2. Create the Dependent Variable: NCD Status
* Assuming qsn3_09 codes 1-14 are specific illnesses and some code (like 0 or 95) is "None"
gen ncd_status = 0
replace ncd_status = 1 if qsn3_09 > 0 & qsn3_09 < 15
label define ncd_lab 0 "No Chronic Illness" 1 "Has Chronic Illness"
label values ncd_status ncd_lab

* 3. Clean Education Variable
gen edu_level = 0
replace edu_level = 1 if qsn2_13 == 1  // Primary
replace edu_level = 2 if qsn2_13 == 2  // Secondary
replace edu_level = 3 if qsn2_13 == 3  // University/Higher
label define edu_lab 0 "None" 1 "Primary" 2 "Secondary" 3 "Tertiary"
label values edu_level edu_lab

* 4. Create Agriculture Engagement Variable
gen is_farmer = 0
replace is_farmer = 1 if lab11 == 1 | lab12 == 1
label var is_farmer "Engaged in Farming/Livestock"

* 5. Create Technology Adoption Index
gen tech_access = 0
replace tech_access = 1 if qsn6_04 == 1 // Has mobile
replace tech_access = 2 if qsn6_04 == 1 & qsn6_08 == 1 // Has mobile + Internet
label define tech_lab 0 "None" 1 "Mobile Only" 2 "Mobile & Internet"
label values tech_access tech_lab

* 6. Handling Wealth/Livestock (Agriculture wealth)
gen total_livestock = _v2670 + _v2671 + _v2672 if !missing(_v2670)

* 7. Clean Lifestyle Controls (Smoking & Qat)
gen risk_behavior = 0
replace risk_behavior = 1 if qsn3_22 == 1 | qsn3_25 == 1
label var risk_behavior "Smokes or Chews Qat"

* 8. Basic Logistic Regression Model (Example)
logit ncd_status i.edu_level i.is_farmer i.tech_access qsn1_05 i.qsn1_03 i.ea_type_n risk_behavior [pweight=wgt_adj_pop]

4. Brief Research Rationale

• Education: Higher education often leads to better health literacy but can also lead to sedentary office jobs, increasing NCD risk.
• Agriculture: In Somalia, those in agriculture (especially nomads) have higher physical activity levels, which may protect against NCDs compared to urban dwellers.
• Technology: Mobile phone and internet access correlate with urban “westernized” lifestyles and processed food consumption, potentially increasing BMI and NCD prevalence.
• The Connection: This study will explore if technological and educational progress in Somalia is creating a “health transition” where NCDs are replacing infectious diseases as the primary health burden.