The calculator predicts time-dependent changes in overweight and obesity prevalence by formulating how populations interact and move between BMI classes. Although the default parameters are set using 1993 United States prevalence data, the model can be applied to other countries or specific regions by changing model parameters. One can change the model parameters by selecting the labeled tabs located at the top of the calculator, entering in the numerical change, and pressing enter. Detailed instructions below.

The calculator predicts time-dependent changes in overweight and obesity prevalence by formulating how populations interact and move between BMI classes. Although the default parameters are set using 1993 United States prevalence data, the model can be applied to other countries or specific regions by changing model parameters. One can change the model parameters by selecting the labeled tabs located at the top of the calculator, entering in the numerical change, and pressing enter (see screenshot of user input tabs below).

These model parameters govern the mechanisms which increase or decrease the prevalence within each BMI classification: overweight, obese, and extremely obese. To see a flowchart describing these mechanisms, **click here**.

The calculator allows the user to change the magnitude of these contributing factors and observe the impact on obesity prevalence projections.

In each of the links below we describe the meaning of each one of these parameters and their reported range. With a few exceptions, the term “rate” is used to quantify the proportion of the population per year that transitions from one classification to another classification (e.g., from overweight to obese). The first exception is birth rates, which represent the number of live births per 1,000 people per year in the population. The second are the social influence transition rates which reflect the probability of effective contact between two populations in a BMI classification.

**Percent initial prevalence**

**Birth and death rates**

**Rates that affect the proportion of healthy weight individuals in the population (BMI < 25) and rates of the factors that predispose them to weight gain.**

**Rates that affect the proportion of overweight individuals in the population (25 ≤BMI<30)**

**Rates that affect the proportion of obese individuals in the population (30 ≤ BMI <40)**

**Rates that affect the proportion of extremely obese individuals in the population (40 ≤ BMI)**

**Rates that affect the proportion of overweight individuals that transition back to healthy weight**

Initial prevalence represents the percent of the population classified as overweight, obese, and extremely obese at the beginning of your simulation. The default is set as the Centers for Disease Control (CDC) reported prevalence values for the United States (US) in 1993, but can be changed to known rates from any year and country desired.

Values must be entered as a percent, i.e., a number between 0 and 100. The sum of the percentages cannot exceed 100%.

To have values reflect a different population, we recommend applying nationally reported values. These values are often reported through national health surveys like the Centers for Disease Control’s Behavioral Risk Factor Surveillance System: http://www.cdc.gov/obesity/data/adult.html/.

Estimates of birth and death rates for the US and other countries can be obtained from the Central Intelligence Agency factbook: https://www.cia.gov/library/publications/the-world-factbook/rankorder/2054rank.html.

To determine the proportion of women of reproductive age classified as overweight or obese, we relied on national reported prevalence rates by age (see, for example, the US Centers for Disease Control or the UK national Health Survey).

The mathematical model is dynamic. A dynamic model is typically formulated using differential equations and the estimation is determined by the “inputs” and “outputs” at the different nodes. This type of model differs from a statistically based model which estimates trends by fitting linear or nonlinear curves to existing data. In contrast, a dynamic model formulates the mechanisms that raise or lower the population in each BMI classification. The resulting simulated trend does not follow a pre-designated curve. Rather the curve shape is a consequence of the identified relationships. Therefore, to construct a dynamic model, one needs to first understand relationships between variables which are to be predicted. The following flowchart describes the relationships for which equations were developed for use within the obesity prevalence prediction calculator.

The java applet was designed by Carl Bredlau. © Copyright 2011