For activity measures, animals are routinely tested in an open field. The historically used apparatus for this is a 182.9 cm by 182.9 cm box constructed from plywood painted black with the bottom composed of 16 equal size squares. However, open field measurements can now be recorded using the more advanced computerized open field activity chamber, consisting of a Plexiglas box (43.2  ×  43.2 cm) with a white floor, opaque walls, and infrared photocells which detect both vertical and horizontal movement (MED Associates, St. Albans, VT, USA). The latter approach provides a more precise analysis of locomotor behavior. With the use of other parameters such as vertical counts, distance traveled, average velocity, and peripheral versus center movement, the computerized open field activity chamber can more accurately detect additional behaviors related to exploration and anxiety.

For determination of circulating lipid levels in rat serum samples, an enzyme kit from Sigma-Aldrich (St. Louis, MO) is commonly used. The Serum Triglyceride Determination Kit utilizes a Free Glycerol Reagent, Triglyceride Reagent, and a Glycerol Standard. The assay involves enzymatic hydrolysis of serum triglycerides to glycerol and fatty acids. The glycerol undergoes several enzymatic reactions, resulting in a quantifiable color change that is directly proportional to the triglyceride concentration of the sample. The color change is detected using an Emax microplate reader (Molecular Devices, Sunnyvale, CA, USA) set at a wave length of 540 nm.

Several important studies examining patterns of fat consumption have demonstrated that initial fat preference and weight gain on a high-fat diet have a strong relationship to long-term consumption of fat, body weight gain, and body fat accrual (23–25). Building on this naturally occurring phenomenon, subsequent publications have successfully utilized measures of early fat preference and initial weight gain on this diet to identify animals with a greater propensity to overconsume a high-fat diet over the long term as well as gain more weight and accumulate heavier fat deposits (9, 10, 16, 26). These studies performed in rats have demonstrated that animals preferring fat over other macronutrients, such as carbohydrates or proteins, are more likely to eat greater amounts and gain weight on this diet. Likewise, rats that gain the most weight over the first 1 or 2 weeks of high-fat diet exposure have an increased propensity to develop an obese-like state when allowed to consume this diet for several months.

In order to have accurate measures of feeding behavior, factors affecting the animals’ stress levels should be kept to a minimum, and measurements should be consistent from day to day. Animals should be maintained on a 12 h light cycle, preferably with the dark or active period occurring during the evening when there are few people in the laboratory. They should be acclimated to standard housing conditions, handled daily for at least 1 week prior to diet exposure, and acclimated to the diets themselves prior to ad libitum feeding in order to avoid effects of neophobia. To prevent the diet from spoiling or becoming less palatable, it should be prepared fresh every week and kept refrigerated until served, and the animals should receive fresh diet every 2–3 days. Measurements of food intake should be recorded daily at the same time each day, preferably immediately after the end of the dark cycle.

In more recently published articles, a modified version of the fat preference or weight gain model has been used to more easily and accurately classify rats based solely on their initial consumption of a fat-rich diet (12, 13). Using this protocol, animals can be subgrouped into high-fat consumers (HFC) and Controls based on their initial intake of a high-fat diet consisting of 50% fat. Specifically, after the 1-week acclimation period to laboratory conditions, chow intake should be monitored for 3 days and then the high-fat diet introduced. This introduction to the new diet should occur over 3 consecutive days, with a 15 kcal high-fat meal given to the animals along with their daily chow. By the end of the third day of exposure, all subjects should have learned to consume the entire high-fat meal by consuming it at least once. After this acclimation, chow is removed, and the animals are allowed to consume the high-fat diet ad libitum for 5 days, with measures of fat intake and body weight taken daily. Animals are then rank ordered based on their fat intake, with the top third forming the HFC group, which consumes about 35% more daily calories with the high-fat diet rather than chow, and the bottom third forming the Control rats, which tend to consume equal calories from both diets. Using this classification procedure, HFC rats have been shown to be more prone to consuming excessive fat over the long term, both during chronic high-fat diet access as well as during reexposure after a 2-week withdrawal from this diet (12). Further, when maintained on the diet chronically, these HFC rats ultimately gain more weight and develop larger fat pads (12). This model is also validated by other studies showing similar measures of energy intake and weight gain to be accurate indicators of long-term patterns of fat consumption and mild obesity (9, 11, 16, 26, 27).

Locomotor activity has been closely associated with food seeking and is particularly high immediately prior to scheduled palatable meals (28). Several studies have related high activity levels to the consumption of palatable foods and other reinforcing substances (13, 29, 30), suggesting that locomotor activity may serve as a good indicator of future patterns of fat intake. The procedures commonly employed to classify animals by their locomotor activity are described here.

Locomotor activity is usually tested within specialized open field chambers (described above), in which either an observer blind to the study or a computerized program records behaviors related to horizontal and vertical motion. Testing should be carried out during the animals’ waking time (i.e., the dark cycle), as this is when their baseline activity is highest (31). Depending on whether animals have been acclimated to these chambers or are experiencing them for the first time, the behavioral analysis can be used to determine pure locomotor activity or novelty-induced locomotor activity, respectively. Several published articles have suggested that novelty-induced activity is a more reliable predictor of subsequent intake of reinforcing substances, including fat (13, 32, 33). Another important factor to consider for novelty-induced locomotor activity is the duration of the test. Optimally, activity measures within the first 5–15 min of exposure to the arena are examined, a time when the novelty of the environment appears to have the greatest impact (31, 34). The specific measurements frequently used to represent locomotor activity are either line crossings when scored manually or ambulatory distance in the case of computerized programs. For novelty seeking, an additional measure of rearing behavior (manual) or vertical counts (computerized) can be recorded.

Using these measures, SD rats have been successfully characterized as HFC and Controls based on their levels of novelty-induced locomotor activity (13). With this protocol, each rat prior to any fat exposure and in the middle of the dark cycle is placed in the center of the open field, and the number of lines crossed is recorded for a minimum of 5 min, with the placing of both front paws and torso into a new square counted as a line crossing. Between tests, the apparatus needs to be thoroughly cleaned with 70% EtOH and allowed to dry. After the activity testing, animals can be given the high-fat diet for classification into HFC and Controls as described above. These initial measurements of line crossings have been found to be greater in the HFC rats that are prone to overconsuming fat over the long-term, suggesting that this measure of novelty-induced locomotor activity may itself be a strong predictor of future fat consumption. In recent studies, this same measure of novelty-induced activity has also been found to strongly and positively correlate with ethanol consumption, suggesting that this behavioral trait may be important for identifying animals prone to over consuming drugs of abuse as well as fat (13, 35).

Circulating lipid levels such as serum TG, which are elevated by the consumption of a high-fat diet (14, 36), are a good indicator of metabolic activity and efficiency. In the body, these lipid molecules can either be stored in adipose tissue for future use or broken down for immediate energy in the form of fatty acids that are known to produce a feeling of satiety (37). When they accumulate in serum, however, they may signify inefficient metabolism and a disruption in energy homeostasis (38). Based on this information, it is likely that animals showing exaggerated exaggerated TG levels after a fatty meal may find this diet less satiating and therefore go on to overconsume the diet. This principle has recently been supported in animal studies showing that, compared to a low-fat meal, a meal high in fat content that markedly elevates TG levels also leads to hyperphagia in a subsequent test meal (39). Based on this idea that exaggerated fat-induced TG levels may be an indicator of metabolic inefficiency and reduced feeding satiety, it has been suggested that these lipid levels may serve as a valid predictor of future fat overconsumption as well as obesity.

According to a recent publication, measurements of fat meal-induced TG levels can be used to predict increased caloric intake and dietary obesity in rats (14). In order to use TG levels to identify these animals, SD rats are first acclimated to standard housing conditions and trained to consume a high-fat meal in a manner similar to the high-fat acclimation for ad libitum consumption. The meal is 15 kcal of a 50% high-fat diet (see Table 1), given for 30 min each day over 3–4 days until all subjects have consumed the entire high-fat meal at least once. This should occur at dark onset since animals are more likely to consume this meal during their early waking hours. After this training, animals are tested for their TG response to fat by exposing them to a similar small 15 kcal meal of high-fat diet once daily over 3 nonconsecutive days and collecting their tail vein blood 1 h after each exposure for measurements of serum TG levels using the Serum Triglyceride kit described above. During these tests, chow is removed prior to dark onset in order to prevent nonspecific food intake, which could interfere with TG measures, and also to motivate animals to consume their entire test meal prior to blood sampling. On the basis of their fat meal-induced TG levels, animals can then be rank ordered, with the top third representing the predicted fat overconsumers and the bottom third being their lower fat-eating (Control) counterparts, with the middle group omitted.

Using this approach, a high-TG response to fat can reliably identify animals with a greater propensity to consume excess amounts of fat during chronic access and as a result to exhibit certain metabolic disturbances. This model differs from the 5-day fat intake and activity models described above, as it classifies animals based on their TG response to three test meals of a high-fat diet, which is a metabolic rather than behavioral parameter. Although the animals characterized as high TG responders consume similar amounts of daily high-fat diet (117  ±  2.6 kcal) as the HFC animals described above (124  ±  3.5 kcal), it is not clear whether they in fact represent the same subgroups of animals. However, with TG levels also increased in the HFC rats predicted based on their 5-day fat intake, it is likely that animals predicted based on their initial consummatory patterns and TG response to fat are indeed the same subgroup of prone rats. Other support for the TG prediction model comes from studies demonstrating that elevated fasting TG levels can predict future weight gain (40) and that lipid-lowering drugs reduce food intake in obesity-prone rats (41).

When using 5-day fat consumption to identify HFC rats, there are two main variables to monitor, namely, diet consistency and day-to-day consumption patterns. Since this prediction is based on the animals’ consumption of this diet over a short 5-day period, it is important to keep all ingredients and the texture of the diet identical for each animal. When a 50% high-fat diet is used, this diet should have a completely smooth texture, while other lower fat diets may have a more powdered consistency. The smooth texture allows the diet to be made into a small 15 kcal ball and placed right in the rat cage for training purposes, and it is also important for keeping the diet in a metal bowl or glass jar during periods of ad libitum feeding, rather than having it spilled by the animals. Unaccounted-for spillage will surely lead to incorrect rank ordering of animals and thus unreliable results. Additionally, it is important to keep the fat concentration at 50% since animals seem to find this amount highly palatable and therefore overconsume it in the short 5-day time period. Whereas a higher-fat diet of 60%, shown previously to predict long-term fat consumption and weight gain, may also be used (9, 26), more recent studies have found a 50% fat diet, which is closer to the human diet, to yield reliable and replicable results (12, 13). Aside from the fat consistency, it is important to carefully examine and correlate the rats’ fat consumption each day. The animals should only be subgrouped once their day-to-day consumption is stable, as indicated by a positive daily correlation of >0.60 as in previous publications (12, 13). The lack of a stable consumption pattern could signify a problem with either the animal’s health or the diet consistency and therefore should be investigated prior to classifying animals as prone or resistant.

When using novelty-induced locomotor activity to classify animals as fat overconsumers, measurements of activity should be made in a uniform and standardized manner, and careful attention should be paid to the animals’ stress level. Standardizing the measurements is especially important for manual recordings, as some behaviors may be hard to capture and track with the human eye. For example, line crossing should be counted only if the animal has both paws and torso in a new square, and rearing behavior should be counted if the animal is fully on his hind paws for a minimum of 2 seconds. If the observer has any doubts about these measures, use of a camcorder may be advisable to videotape the test session and then analyze the behavior later based on the recording. Also, as animals may be anxious upon exposure to a new environment, it is important to note any particular anxiety measures in the open field. One measure of such behavior is the amount of time spent in the periphery versus the middle of the open field, with greater peripheral time representing an anxious state. Since anxiety can mask increases in novelty-induced locomotor behavior, it is important that the animals are well handled and gently placed into the open field apparatus for testing.