Research Title

Factors Contributing to Obesity Among Adults: A Statistical Analysis Using Minitab

By Bonaventure

Problem Statement

Obesity has become one of the most significant public health challenges worldwide. It is associated with numerous health conditions, including Type 2 Diabetes, Hypertension, cardiovascular diseases, and certain cancers.

Modern lifestyles characterized by reduced physical activity, excessive calorie consumption, poor sleeping habits, and sedentary behavior have contributed to increasing obesity rates among adults.

Understanding the factors associated with obesity can help healthcare professionals and policymakers design effective intervention programs.

The prevalence of obesity among adults continues to increase despite growing awareness of healthy lifestyles.

There is insufficient statistical evidence regarding the relative contribution of age, exercise frequency, calorie intake, and sleep duration to obesity among adults.

This study seeks to identify the major factors influencing obesity and develop a predictive model for Body Mass Index (BMI).

General Objective

To determine the factors contributing to obesity among adults.

1. Determine the average BMI among adults.
2. Examine the relationship between age and BMI.
3. Assess the effect of exercise frequency on BMI.
4. Investigate the influence of daily calorie intake on BMI.
5. Determine whether sleep duration affects BMI.
6. Develop a regression model for predicting BMI.

Research Questions

1. What is the average BMI of adults?
2. Does age significantly affect BMI?
3. Does exercise frequency influence BMI?
4. Does daily calorie intake influence BMI?
5. Does sleep duration affect BMI?
6. Which factor contributes most to obesity?

Research Hypotheses

Hypothesis 1

H₀₁: Age has no significant relationship with BMI.

H₁₁: Age has a significant relationship with BMI.

Hypothesis 2

H₀₂: Exercise frequency has no significant effect on BMI.

H₁₂: Exercise frequency significantly affects BMI.

Hypothesis 3

H₀₃: Daily calorie intake has no significant effect on BMI.

H₁₃: Daily calorie intake significantly affects BMI.

Hypothesis 4

H₀₄: Sleep duration has no significant effect on BMI.

H₁₄: Sleep duration significantly affects BMI.

Data Collection Method

A questionnaire or survey was used to collect data from respondents in Huston, Texas. Minimum: 250 respondents.

Key Findings: Insights

Press enter or click to view image in full size

descriptive statistical summary

Insights from the Data

• Sample Size: 250 people — reasonably large and solid for this kind of analysis.
• BMI: Average is 23.6 (normal weight range). The distribution is fairly typical, with most people between 21.7 and 25.4. A few individuals reach obese levels (up to 34.8).
• Calorie Intake: Average ~2032 kcal/day. Quite a wide spread (±358), which explains why it showed the strongest effect in the Pareto chart.
• Exercise: People exercise on average ~2.5 hours/day. This seems relatively high — possibly includes walking or general activity rather than intense gym time.
• Sleep: Very normal distribution centered around 6.8 hours — typical for adults.
• Age: Well spread from 18 to 64, centered at 41. Good variation for analysis.

Outlier Detection in BMI

Press enter or click to view image in full size

The BMI values in this graph above are mostly concentrated in the normal/healthy range (21.7–25.4), with a few people having higher values that appear as outliers on the plot. The average and median are nearly the same, indicating a fairly balanced distribution with a slight positive skew.

Press enter or click to view image in full size

Both visually and statistically, the Normality test chart shows that “the BMI data is normally distributed”. We can confidently use parametric statistical methods (like t-tests or ANOVA) that assume normality on this dataset.

Correlation Test of Features

Press enter or click to view image in full size

Statistically Significant Relationships

Three relationships in this data stand out as statistically significant, all of them involving BMI:

• Daily Calorie Intake vs. BMI ($r = 0.371$, $CI: [0.258, 0.473]$)
• Interpretation: This is the strongest correlation in the dataset. It is a moderate positive correlation, meaning that higher daily calorie intake is significantly associated with an increased BMI.
• Daily Exercise vs. BMI ($r = -0.246$, $CI: [-0.359, -0.126]$)
• Interpretation: This is a weak-to-moderate negative correlation. The negative value means that as daily exercise increases, BMI tends to decrease, which aligns with typical health expectations.
• Age vs. BMI ($r = 0.128$, $CI: [0.004, 0.248]$)
• Interpretation: This is a very weak positive correlation. Because the lower bound of the CI (0.004) is just barely above zero, there is a statistically significant trend where BMI increases slightly with age, though the effect is very small.

Non-Significant Relationships

For all other pairs of variables, the 95% Confidence Interval spans across zero (meaning it includes negative numbers, zero, and positive numbers). This indicates no statistically significant linear relationship between them in this sample.

Most notably:

• Sleep Duration vs. BMI ($r = -0.098$, $CI: [-0.220, 0.026]$): While the negative correlation suggests a slight trend where less sleep equals a higher BMI, it crosses zero, so we cannot claim a definitive linear relationship.
• Age vs. Lifestyle Choices: Age does not have a significant correlation with Daily Calorie Intake ($r = 0.031$), Daily Exercise ($r = 0.112$), or Sleep Duration ($r = -0.016$).

Multi-Regression Analysis Using Pareto Charts

Press enter or click to view image in full size

According to this analysis, Daily Calorie Intake is the most influential factor on BMI, followed by Exercise and Age. Sleep duration appears to have little to no detectable effect.

Conclusion and Recommendation

Conclusion

This study evaluated the impact of demographic and lifestyle factors specifically Age, Daily Calorie Intake, Daily Exercise, and Sleep Duration on the Body Mass Index (BMI) of 250 respondents in Houston, Texas. Based on the rigorous statistical analysis, the hypotheses are resolved as follows:

• Hypothesis 1 (Age vs. BMI): Reject $H_{01}$ There is a statistically significant, though very weak, positive relationship between age and BMI ($r = 0.128$, $CI: [0.004, 0.248]$). BMI tends to slightly increase as individuals age.
• Hypothesis 2 (Exercise vs. BMI): Reject $H_{02}$ Daily exercise frequency has a statistically significant, weak-to-moderate negative effect on BMI ($r = -0.246$, $CI: [-0.359, -0.126]$). More exercise is reliably linked to lower BMI.
• Hypothesis 3 (Calorie Intake vs. BMI): Reject $H_{03}$ Daily calorie intake has the most substantial, statistically significant positive effect on BMI ($r = 0.371$, $CI: [0.258, 0.473]$). This was further verified by the Pareto chart as the leading predictor of weight variance.
• Hypothesis 4 (Sleep Duration vs. BMI): Fail to Reject $H_{04}$ Sleep duration has no statistically significant linear effect on BMI ($r = -0.098$, $CI: [-0.220, 0.026]$). The confidence interval crosses zero, indicating that sleep variation does not predictably influence BMI in this sample.

In summary, personal lifestyle choices chiefly caloric intake followed closely by physical activity levels exert the strongest mathematical influence on an individual’s BMI, overriding uncontrollable biological components like chronological aging.

Recommendations

1. Target Caloric Regulation as the Primary Intervention

Since daily calorie intake emerged as the most dominant factor in the Pareto analysis and correlation tests, public health campaigns, coaching programs, or corporate wellness initiatives should prioritize nutritional literacy. Focus efforts on caloric mindfulness, portion control, and tracking food intake, as this metrics yields the highest statistical leverage for managing BMI.

2. Promote Accessible Daily Activity Over Intense Gym Regimens

Given that the sample reported a high average of ~2.5 hours of daily activity (likely incorporating light movement, walking, and occupational steps) and showed a significant negative correlation with BMI, individuals should be encouraged to build sustainable active habits. Focus on cumulative daily movement (e.g., hitting a daily step goal) rather than just structured, high-intensity gym sessions.

3. Account for Age-Related Metabolic Slowdowns

While the correlation between age and BMI is weak, it remains statistically significant. Health guidelines should include specialized advice for aging populations, highlighting that nutritional requirements and metabolic rates shift over time, requiring minor adjustments in intake to prevent gradual weight gain.

4. Contextualize Sleep in Broader Well-being Frameworks

While sleep duration did not demonstrate a direct linear effect on BMI within this specific sample, it remains essential to overall physiological health. Future iterations of this survey might benefit from tracking sleep quality or sleep disruption metrics rather than just total hours, as quality often correlates more accurately with metabolic health than duration alone.